CZ and Friends
CZ & Friends is a podcast about what it takes to lead and evolve legal in an era of exponential change. Hosted by Cecilia Ziniti, former General Counsel turned founder and CEO of GC AI, each episode features candid conversations with legal and business leaders who are building for scale, taking bold bets on technology, and leading with humanity. Whether you're a GC, operator, or in-house counsel, this podcast is your front-row seat to the future of legal.
CZ and Friends
Instacart Legal: Making AI Everyday Practice
Today’s guests are Nicole Altman, Senior Counsel, and Kelly Noguchi, Senior Legal Technology and Operations Manager, at Instacart. They are helping shape how a public company’s legal team experiments with, adopts, and scales AI. Together, they show what innovation looks like when it is encouraged and operationalized.
You’ll hear:
– How Instacart Legal made AI exploration part of team and company-wide goals
– Why curiosity and a learning mindset are the foundation for adoption
– Real examples of tools like Contract Compass and Legal@ triage in action
– How leadership and culture drive not only productivity but creativity
– Why legal ops is critical for moving from early experiments to enterprise systems
Nicole and Kelly reveal the human side of scaling legal with AI: the joy, the overwhelm, and the opportunities to lead change across the business.
Follow Nicole & Kelly:
@Nicole Altman on LinkedIn
@Kelly Noguchi on LinkedIn
Books, Authors & Thinkers Mentioned:
– The Worlds I See by Fei-Fei Li
– Radiolab podcast
– More Perfect (Radiolab’s legal spinoff)
Other References:
– MIT study on AI adoption and productivity
– Instacart’s “AI First” company initiative
– Bolt hackathon (Contract Compass project)
– Enterprise AI tools and internal chatbot (“Ava”)
Follow us on all social platforms to get each new episode when it drops.
@Cecilia Ziniti on LinkedIn
@CeciliaZin on Twitter/X
@GC AI on LinkedIn
gc.ai website
It is getting so good that it doesn't replace us. You know, I think, you know, there's a lot of talk about, you know, is AI going to replace us? And I I personally don't have that fear, but it can be a really, really amazing companion and assistant and thought partner in how we do our work.
Cecilia Ziniti:Welcome back to CZ and Friends, where we explore how legal leaders are shaping the future of business and law and beyond. I'm your host, Cecilia Ziniti. Today I'm joined by two leaders from Instacart who are turning AI and legal from hype to impact. These two women are incredible legal team members, and I'm so excited to speak with them. Nicole Altman is senior counsel at Instacart, where she leads AI governance, privacy, security, and IP. She's built AI tools that have won hackathons. She's been recognized in Instacart's earnings call, and she's served as the head of Women at Instacart, the company's largest ERG. Kelly Noguchi is senior legal technology and operations manager at Instacart, where she drives adoption of new technologies, builds systems that improve legal workflows, and leads efforts to make AI more accessible and effective across the team. Let's dive right in. Nicole, Kelly, welcome to the show.
Nicole Altman:Thank you so much, Cecilia . So happy to be here.
Cecilia Ziniti:Same here, so excited. Awesome. Let's dive in. All right, so it's 2025. Everyone is saying this is the year of AI agents. We've got legal teams are some legal teams are experimenting, some are a more forward-thinking. You two, I understand, are really at the forefront. Venture backed company moving fast. You've embedded AI. Say that one more time.
Nicole Altman:Public public company. Is that when it was?
Kelly Noguchi:Like I feel like it was right after the pandemic. We were the first one, like the first post-pandemic idea. All right. Well, so I stand corrected.
Cecilia Ziniti:So originally venture back company, but in any event, operating in many ways like a massive startup, adding new geographies, serving millions of customers. So I think we found here you all you all are different. So you've embedded AI into workflows and you're seeing results at Instacart Legal. Tell us about that.
Nicole Altman:Yeah, you know, I I feel very fortunate to be at Instacart. I joined at the end of 2021 when uh generative AI wasn't on my mind, wasn't on the public, public conscious. But Instacart is very AI forward, I think, you know, even among tech companies. And I feel really grateful to be at a company like that where AI exploration and experimentation is encouraged, even mandated. And we're giving like the the time and space to do that. And for someone like me who finds it endlessly exciting and fascinating, it's just wonderful. I love that. How about you, Kelly? What's your take?
Kelly Noguchi:Yeah, I mean, generally I like I feel so fortunate. Like even if it's overwhelming sometimes, and I could definitely get into that too. Like I when I have conversations with other people, like I realize how much like just advancement we got thanks to being at Instacart, just having the opportunities and support and just having that embedded in our day to day, it's it makes a huge difference for sure.
Cecilia Ziniti:So when you say AI is embedded in day-to-day, or that innovation spirit is embedded in day to day, paint paint us a picture. So, like, is there like Instacart leadership principles or like what is it about the company that makes it as you described, Kelly?
Kelly Noguchi:Um, I mean, like Nicole mentioned, it's I would say like softly mandated, right? Like it's part of our our goals, like our personal team goals, company-wide goals. So everything's like our um, it's oriented towards okay, being fluent as fluent as possible with AI and how we approach that, like they give us a lot of leeway. So in that sense, like I feel like um, you know, they just are able to say, okay, AI is gonna be a it's it's something that we all have to learn no matter what, and how you learn it and how you incorporate into your day-to-day, like it's up to you. But that actually like I think engages a lot more people. So, you know, it is like just company-wide initiatives and then also, you know, understanding that, you know, personal productivity is tied to AI too. So like learning from all of the the small tools that we have that's available to us towards um building like very scalable operations with AI. It's I feel like AI is just part of every conversation we have of that.
Nicole Altman:Yeah, we were, I mean, you hear now, still now, like you know, three years into the gender of AI and the public consciousness explosion, uh, you still hear of companies that don't have enterprise AI tools where people are using shadow eye because the company is not there yet. We're the opposite. We got enterprise AI like very early on. And so, you know, our our engineers built this internal chat bot built on APIs of these companies that has been accessible to us from years ago, company wide. And so it's been fantastic. We've had the ability to get hands-on early on, and you know, adoption and use is like growing and growing. Um, and now you know there's there's more and more things that we're able to do every day, every week, it seems.
Cecilia Ziniti:A lot of our listeners are, you know, working on AI themselves or thinking about how to do it. Um, do you have specific examples of things that you do and then you know, maybe walk us back to, you know, three years ago, two and a half years ago, whenever you started to do these things, how did you kind of kick them off?
Nicole Altman:There's two broad categories of uses in my mind. There's the day-to-day simple interactions with a chat bot that can have these very meaningful productivity and creativity boosters. That is very easy, very accessible to do. And for most, if not, I mean, probably upwards of 90% of my work, I am consulting with an AI chat bot to refine my ideas, to help brainstorm, to be a thought partner, to critique my email, to critique my strategy, things like that. And then there are the bigger types of systems that it takes a little bit of time and investment to build these workflows. And there's there's a great time and place for both. But I think you mentioned at the start, you know, agents are like the shiny new thing. Everybody's talking about agents. And that's great and good, but there's so much amazing things that AI can do just with the the chatbots. We have these monthly um MBRs where people from different uh pillars within the team present on what they're doing. And just yes, was it yesterday or the day before, Kelly? Somebody had a slide like iHeart Chat GPT talking about how it's just you know radically transformed her work and the way she goes about it because it is getting so, so good that it doesn't replace us. You know, I think you know, there's a lot of talk about, you know, is AI going to replace us? And I I personally don't have that fear, but it can be a really, really amazing companion and assistant and thought partner and um partner in how we do our work.
Cecilia Ziniti:So you're in this this MBR and one of your attorneys on the team is so excited about the use of AI. What did you as a legal ops person do to get them to that point, if anything? And you know, it is part of the culture, but like there, there's clearly like in legal ops, people talk about driving adoption, people talk about getting people over that hump of fear. You know, was that something deliberate on your side? Obviously, you're super experienced. Give us a little bit of the legal ops view.
Kelly Noguchi:Yeah, I mean, weirdly, I always like have, well, nowadays I have like mixed feelings about people's reactions to the use of AI. And like, so to like try to answer your first your question about like adoption of even like the basic chat. So what you know the user was excited about was like the chat function, the personal product like productivity chat mostly. So being able to engage with chat to uh with the chat bot like Chat GPT to be able to come up with better, you know, outlines or doing helping with research and things like that. And that's like so important. And I feel like people like a lot of studies, like including the MIT study that came up, like ignore that kind of like like boost for personal productivity because that's huge. And you know, like Nicole said, I don't know what you know, I did before Chat GPT. Like I feel like almost every part of my task day-to-day, personal too, like I'm engaging with Chat GPT or you know, some sort of art. Like, um, and so I think I'm gonna segue a little bit into the MIT stuff because it's so top of mind for me. And like as a legal ops person, if you know, we're talking about like implementation, like um that gap between personal productivity to a scalable, like business ready, enterprise ready process is where that huge challenge is. And it's highlighted by the MIT study, which I know you could like totally poke holes in. And I love people to like poke holes because I I do see all of the issues with it, but um, like I think the sentiment is real though. And I'm like on social media, that's like where I usually like kind of learn about um AI, like AI trends and what's going on and how people are engaging with AI. Like, you know, I see both sides of like I feel like it's kind of two main sides. Like, is AI overhyped? Like this MIT study proves it, or two, like, you know, that's like this is BS, like, you know, there the like AI is like is is like a a game changer, which I agree. And I feel like both are are right. Like the emotional impact of the study is like true, right? Like if you're in social media, if you've been on TikTok, you'll see it. Like the people that are trying to actually implement AI, it's you know, there's challenges, like any change management that goes on, but AI, like just the way the AI was involving, the pace, just just the massive influx of different tools. And I feel like each model is like being like pushed out every two weeks. Like that that actually like makes like tool implementation just completely different from what you know, legacy, like I what you know the folks are used to, like with regular staff tools. Like I feel like people had a really structured cadence. There's usually like certain tools that were, you know, like industry standards in the market, and you know how to do an RFP. It's like even roadmaps for these tools are pretty structured too. And so you could do these long-term plans, have these deep RFPs, like it's just not the same anymore with these AI tools. And so I could see like there's like even that RFP portion, RFP portion that's challenging, and then the implementation. Like, I feel like we are expected, like Nicole, like you like I feel like I asked this of the team constantly to be like basically testers, QA people on top of their day jobs all the time, right? Like they have to continuously learn, continuously be creative. Like it is like it, you know, it's like a real thing. So I could imagine that's why I like the MIT, feel like it like both sides of the argument blew up a little, and like I could I could as like kind of person in the middle, I a hundred percent like understand. Like I'm so like I'm an AI enthusiast for sure, but also like I think we can't like glaze over the actual like changing like excitement into execution. Like that's you know, that's real.
Cecilia Ziniti:Yeah, so love that. So I feel like you and the MIT study and Nicole were kind of converging on this idea that this big kind of like, you know, we're gonna revolutionize overnight and process X is gonna go become process Y immediately. It's this huge thing, end-to-end agents, etc. That's kind of one category of AI, but the category of AI that you're having success with and that that you know Cole pointed out is that creativity, that initial, like I'm using it for my job in legal. I'm doing a data privacy impact assessment. I'm gonna have AI help me specifically go through a checklist, or I'm gonna have AI help me, you know, review a contract or what have you. And so that aspect, I guess you you homed in on something super interesting, Kelly, which is this like creativity, curiosity kind of that you're asking of Nicole. So switching it over to Nicole, like, is that a like a personality trait? How do you like when I think about our users? There is an adoption curve, right? So we're gonna drop the MIT study in the show notes so you can catch that in the show notes. But the MIT study basically said at a lot of companies, there's these top-down AI implementations and they're not working as expected because individual folks are not being supported as much. That's the thesis. You're the counterpoint, it seems, Nicole. Tell us about you and like I I want to get real like, what makes you an early adopter? And when Kelly sends you an email, she's like, you gotta test tool XYZ. What makes you not hit delete?
Nicole Altman:It's just it's it's fascinating. I I love it. It's it's magical what it can do. These tools now open up these possibilities that were not possible before. But it it takes time, it takes a lot of investment of time and grit and patience to get stuff out of them. You know, early on I was talking to some friends, not Laurie's, just about who's using it. And well, you know, I I tried it, but I I I tried to write a condolence card for someone. It was too flowery. It's like, did you try to ask it to make it less flowery, f flowery? You can do that. All right. It's a like, you know, that's a very small, simple example, but it like you do need to like push through and really get hands-on to build that intuition of what the system can do and cannot do. And you have to keep doing that over and over again because the systems are so rapidly changing. It's not static. So what it can and cannot do today uh is very, very different than what it can and cannot do three months ago, one month ago, one year ago. And so you have to like it, it does take a lot of investment of time and just being interested in it. And there are times where I get really busy with my traditional job. I'm still using it. Yeah, right? Well, I I didn't want to say that because I I do think that AI is it is part of the job. I get it. I'm still using AI, but I'm not, you know, when I don't have the time and space in my life to like be learning and exploring and pushing the boundaries. I like if that happened for a week or two, I feel so behind. And so like Kelly mentioned, it's overwhelming. And I think just like the pace of it changing and expanding is is overwhelming, and it takes a lot to to get into it and keep up. Um you know, you asked also like what what we've done to get people on board, and it's a it's a long process, like change management is a long process, but I really think that the best way to get people into using AI is to show, show concrete examples of what it can do and how to do it and walk people through it. And so Kelly and I and others on the legal team have had a lot of these sessions for people on, you know, for our team to show, hey, this is you know, step one, this is how you set it up, this is how you access it. Then like these are the things you can do. This is what I do, and just showing that and not just once, but periodically, and you know, like leveling up a little bit each time so it's not boring, but and then like showcasing other cool use cases from across our team, putting out polls, who's doing what. Let's present that to the team. I personally learn so much by seeing how other people are using it, what cool things they've thought of. And so it's just this continual like learning, sharing environment. We have that across the company beyond the legal team as well. We have a Slack channel where where people can put in the cool things that they're doing. And so it it is this continual learning that you need to stay on top of it.
Kelly Noguchi:Like honestly, like we kind of kind of going back to what Nicole already like said in the beginning, it we had, I understand like the top-down like issue that the MIT like study pointed out, but like that actually I think did help with adoption at Instacart, I would say. Like there was, you know, a lot of just constant reminders about using AI and like even the the the basic like chat engagement, like that, you know, like like you said, like it's you know, that's the step, like that's the first step you have to take. And even that's hard for some folks. And but like when you're in a company and everybody else is using it day to day and you have like all these cool like use cases and you're not doing it, like that's you know, I think people realize, ooh, I should check it out.
Cecilia Ziniti:Yeah, I I love what you're saying. I think I think this like you painted a pretty clear picture, the two of you, of like, okay, this is a couple of champions, which it sounds like the two of you, company-wide receptivity to this, and then just a lot of like day-to-day, right? So you're going into this meeting and you're literally showing use cases, you're sharing them, it sounds like in Slack, you're you're saying, okay, I did that. So there um was your general counsel involved, or like what was the because because what I've found, and you know, folks listening, check out the episode with Alice Hunt and I'm sorry, Alice Davidson and Jenna Hunt from Topolity, because they did this at a smaller company and kind of kind of drove it. But very similar themes is what we're talking about here around getting people to to change, getting people to embrace this new way of working, you know, takes some work and you got to kind of also make it fun. So, all right, so there we are with that. I guess question for you is like, what was your general counsel's level of involvement? And then, you know, I have to imagine that even on a very forward-thinking legal team, there's a natural adoption curve, right? So I went into one legal team and they told me, I said, hey, our corporate guy, he's the Eeyore of the group. If you can get him using AI, you will have one. And I'm very happy to report that Eeyore is using GCAI to do his proxy statements and like kind of do some of the research for that. So that was exciting. So two questions wrapped up in was your GC involved? And then how do you handle the kind of back half of the curve, which is naturally you're gonna have? And um, I think you're all our legal department of 60 or so, right? So, so how do you deal with that?
Kelly Noguchi:I mean, Nicole, like I gotta say, like Morgan, our GC was heavily involved. He was very excited from the get-go. And I feel like, you know, I get like the constant existence of AI in our meetings and in our initiatives and in our projects, and just it was just, you know, part of part. I I I can't even, you can't like divorce any of it. Like, I feel like once once we had our um Ava or Instacart's like pro like uh chat chatbot, like I feel like, you know, they made it really easy for us to add it to our calendars or to like put it as a broad browser, like Chrome browser. I feel like, you know, and you know, all of this, like I feel like our GC constantly just reminded us, like this is this is to, you know, this is a like a net positive thing just to check it out. And so, um, and then he built up like all of these committees that you know Nicole mentioned and activities for us to be able to share our use cases and drive adoption more and lift off the folks that might be a little bit less hesitant or like keyed into all of the the chatter that's happening around AI.
Nicole Altman:Um yeah, I mean, case in point, we periodically get together as a team, convene in person, and our last one was uh earlier this year in San Francisco, and the theme of it was cultivating an AI first mindset. And we devoted a significant portion, the majority of our time together, or at least our working time together, to like exploring AI use cases, like bringing up everybody, like rising the tie of everybody to have like shift to an AI-first mindset, seeing what other people are doing, troubleshooting their problems. What do you wish it could do? What are you having problems with? Let's sit down and talk about it and figure out how you can embrace it more and more to unlock productivity gains, unlock things that were not yet possible. Don't just use it as a productivity tool, but like let's think bigger. What like what can we do now that was not possible without these tools? And so just like it's really just creating that space for the conversations and for the experimentation. And absolutely, our GC is is fully into it.
Cecilia Ziniti:Love it. So you said something about kind of thinking bigger, and I love that. So tie that back to me to the work you all do for stakeholders, right? So your product primacy counsel, your legal ops, Kelly, you know, at that time when I when I came up as product and privacy counsel, a lot of my clients were engineers, you know, were heads of product management, that kind of thing. How does your being AI forward tie back to working with them? And I don't know if you've got any wins or anything like that, or or product launches you were involved in, or any any good stories to share, Nicole?
Nicole Altman:I'm gonna pivot a a little bit and talk about this legal at Instacart system that I and some of the people on our litigation operations team built. And this turned a the initial idea was to turn this extremely manual process. Legal at Instacart receives like hundreds of emails a week, ranging from super important to complete garbage and everything in between. And it was it was literally someone's almost full-time job to go through it, figure out what needed to be responded to, figure out like of the things that need to be responded to, like who's handling it, like who like getting it to the right places. And it was very time consuming and a lot of manual effort. We thought this is something that AI should be able to help us with. We had actually tried it last year and it didn't work. But tools evolved, things evolved, we try it again, got it. It's great. But then we don't just stop there. Okay, great. Now we have this triage system. It saves a whole lot of manual effort. It's um doing great at that. What more can we do with that? Well, now we have this system that understands what kinds of things are coming inbound to legal. What insights and intelligence can we get from that? What predictions can we make about trends and things? And so that's that's something that we're building out too that would not have been possible previously. I mean, sure, maybe if you wanted to hire some, you know, several full-time people to like trudge through it and categorize and do regression analysis, sure. But nobody's ever gonna do that. But now at our fingertips, we have these powerful systems that can unlock insights that we never would have uncovered previously.
Cecilia Ziniti:I love that example. So basically, so AI, they've got, you know, a team like Instacart, they're operating in a bunch of markets. They've got physical goods, they've got, you know, this multi-started marketplace, employment issues, privacy. I mean, you can just imagine the gamut, you know, of a now public company doing this. And so literally, y'all had legal at, and anybody in the company could email legal at. I'm I'm just this is not not just not just in the company, like external. Oh, external. Oh, okay. So my you know, my my blue cheese is moldy or whatever, and I'm emailing you. Okay. All right. Well, I will say Instacart service is very good. Um, pro tip if you travel as much as I do, instacart yourself chargers to the hotel from the airport. Done that, instacart yourself a microphone like I've got here. And you, you know, you could do it. Uh, it is just absolutely outstanding. Okay, so you basically have this business intelligence that you didn't have before, it sounds like. And so what are you doing with that? And have the have the stakeholders noticed? Like, have the SLAs gone down, or like what's been the kind of like we can maybe shift the conversation to the ROI word, is like, okay, people think it's math or something, but no, it's like what has been the impact?
Nicole Altman:Yeah, I mean, but we have the triage system built up, and that has had like absolutely ROI reducing processing time tremendously. It is no longer somebody's full-time job. The intelligence system was still being built out, but it's an example of the kinds of things that are able to be unlocked that were not not previously possible. I love that.
Cecilia Ziniti:Okay, so tell us about Contract Compass, besides the great branding. What is it? Um, thank you. Guess who helped me uh with that name? Amazing Kelly. All right, see, she's so humble. She's Chat GPT, right? I'm assuming. Dang, okay, all right, good stuff.
Nicole Altman:Well, I actually I bounced around, as I often do, I bounced around between ChatGPT and Claude for um names. I forget who ultimately came up with with that one. But yes, I mean it's just an example. Using using AI as a brainstorming partner, it wasn't the first name it gave me, but I, you know, give me 50 names. Like, you know, most of them are garbage, but there's gold in them.
Cecilia Ziniti:I love that. You know, it reminds me of um so very early on in AI, we had just launched the product, and a friend of mine who is a plaintiff employment lawyer um was using the product, even though typically we're targeted in-house. He had a deposition and the witness was not being a cooperative on a particular issue. He goes to a break and asks GCAI, he's like, Hey, the witness is not cooperating on this question. Give me 25 ways to get at this thing. And he just exactly like you said, Nicole, he said half of them were garbage. But meanwhile, I had 12 questions that just got to the heart of the issue. I got back, that was the meat of the depot. And he's like, This is the new game, basically, that the whole level is gonna rise. And so I want to ask about that, Kelly. Like with the level rising, you've been in legal ops a long time, you've done you've done a lot of different projects across time. What do you see like the lasting impact being here? Do you do you think people are gonna lawyer harder? Do you think, you know, to speak a little bit to the the fear aspect that you know, Nicole, you said you weren't you weren't scared. Why not?
Kelly Noguchi:I mean, I don't I think like just judging, well, I guess you never know what's gonna happen. I can't make predictions anymore, but like um I never could, I guess. But um, I I feel like there's always gonna be um it's just like the shift of roles and tasks, I think, as you introduce new AI tooling. And then like Nicole's like a great example of a person that's like just baked in AI into their, you know, life and it makes her a better like teammate, a better attorney, a better, just now like vibe coder. Like I feel like it just like you know, you're still there and you're doing all of the things and you're a decision maker and you're you know the pilot, all of that stuff, like but you're just superpowered, charged in different ways. Um but all of that, you know, you caveat with like all the like investment that Nicole mentioned about like double checking and doing your background, like you can't take every you know, especially right now, but um, you know, it just takes a lot of investment, but like I feel like the humans are there and that's like a new part, like role for someone like Nicole and myself too. I I like think like you know, talking about launching like new processes and investing and and like finding new tools and things like that, like one of the things I continued, I feel like I've been struggling with this for a while, and I hope like one like if someone from the audience could reach out to me and tell me like some good ideas, like um, you know, with this landscape changing, with the tools like evolving constantly, like um, you know, like choosing between these niche products that are super cool that are you know being like launched like like every two weeks too. But then also the need, like, you know, we have an environment, we have an ecosystem, we use Google Suite, you know, we have all of these other we're in Salesforce, whatever it might be like. And you know, these native, like these gigantic tools are, you know, launching native AI constantly too, and trying to like really understand like like what is the like what is the best way forward incorporating AI? Like, should we really be asking like users to learn like completely new like UI like all the time? Like should you know, like, or like you know, I I just feel like there's so many questions I have about adopting new tools right now and just like implementing it and how much push we should give to our teammates to be engaged in a process, or maybe it doesn't even make sense to like there they should be a little bit more bifurcation between like the teammates and the folks that are like checking out you know AIs like um like actually the MIT like study really pushed me thinking into like okay, like that study just really shows for me not necessarily like there's a failure and like process even or the tools, like, but it's just like showing the transition point we're in right now, right? Like I feel like we're trying so hard. Clearly, like there's tons of like activity right now and adoption and excitement. And then, but some like you know, we're still figuring out like what the how to best approach like actually like successfully like rolling out um AI. Like I always think like, oh, how did people start using like Word or like the internet or like typing?
Cecilia Ziniti:Yeah, it's fascinating. Yeah, it was slow, right? People are still using tax machines, you know?
Kelly Noguchi:Yeah. It's just but the pressure of that is like I think it was the same. Maybe I need to like we need to talk to folks who like, oh, how was your transition at work from like paper to word?
Nicole Altman:You know, right now I I don't have the statistics offhand, but it it was slow. And in fact, the adoption towards using generative AI is much faster than the transitions have been in the past because of like the pressure though, right?
Kelly Noguchi:Like I feel like there's kind of like this collective what do you think like I feel like it's social media and like just
Nicole Altman:It's like it's there's there's lower costs to entry to like you know the adoption of a computer. You had to like go out and buy a computer, and they were these big machines and they were expensive, and so for a long time they were out of reach for most families. Now almost everyone has a computer, almost everyone has internet access. Um and Chat GPT was available for you know public publicly available for free. And so there's there is um this lower barrier to entry, but still, it's still even though it is exponentially faster than the adoption of other technologies, it's still hard.
Cecilia Ziniti:I mean I I empathize so much with you, Kelly, and and with the sentiment and with Nicole too. Like, so yes, there is precedent in history. If you're interested in this historical topic, Chief Justice Roberts wrote a report on the state of the federal judiciary back in 24 talking about literally the federal judiciary adopting like word processors and that this is the you know, AI is that for our era. And similarly, you know, 1996, yeah, it's crazy. 96, the Illinois State Bar was debating if email was ethical for lawyers to use. Wow. Like that was the debate. They sought public comment about it, and to their credit, they concluded, as we have now with AI, that yes, there is a way to use this ethically, as Kelly described, where you're verifying sources, you know, and you're using it to kind of go to that next level. But but Kelly, in terms of reassurance, I can tell you, so you and Nicole are absolutely at the cutting edge. And I think it's it's almost like a like a Silicon Valley kind of thing, I will say, to feel like you're behind. And I think I think also as lawyers, that happens to us is like this like kind of like almost law school-esque of like, you know, you're kind of looking around, look left, look right. I'm like, where are people doing the the part that at least the solace that I take and and that that works for us is like, you know, that this learning can be fun. And so our job as leaders in the legal team is like to create that environment where Nicole, she's like glowing, talking about using new tools and making sure that you cast you set the framework to cascade that to the team. Yeah, this is so fun. All right, Nicole, you you had a strong reaction on the historical piece. What's the thing?
Nicole Altman:Well, well, I mean, you mentioned fax machines, and like my mind was blown when I clerked um on the second circuit in 2009, nine, 2010. And that is how the judges communicated with each other was by facts. And I'm like, oh my god, like it's 2009. Oh, yeah. 2009, and like I have to call clerks from the other chambers and be like, was that a period or a comma that you put there because it's so blurry with the fax machine that I can't I can't do it. So I'm curious if they use um if they use the.
Cecilia Ziniti:Yeah, no, so I clerked about around the same time period on the Ninth Circuit, and and we had knit memos where you wouldn't red line, instead, you would prepare a memo that's like line, you know, 17, paragraph two, the comma should be a period. And we would literally type a memo that was like, you know, dear esteemed judge, whatever, and send that over. And so now I do think like redlining, let's talk about redlining. So, what is contract compass? Does it have anything to do with redlining? Paint us one more picture. Like, this is just so fun to get this like richness of detail.
Nicole Altman:Contract compass is not is not redlining, it's solving a different problem. But I have to credit, even though Kelly didn't work on it per se, I have to credit Kelly with it in part because this is to the thing of like, you know, there's so much to learn and so much to explore, and sometimes you get busy and you don't. And so vibe coding earlier this year, you know, vibe coding, it was like always on my list. Like, I gotta try this, I gotta try this, but it there was always other things that rose higher on the list. Ellie uh gave a demo of vibe coding. I'm like, okay, this is my impetus. I have to jump in. So I jumped in and I joined this Bolt hackathon. Bolt is one of uh one of the main uh vibe coding tools. And together with some other folks, we built Contract Compass, which solves the problem of, you know, a contract is not a single document. There's the, you know, there's the contract, there's the writer, there's all the amendments. The amendments typically don't restate and amend the whole thing. You have to like stitch them and piece them together. And so, me, when I come in and I'm like, want to understand what the terms are, I go into our database and I search for the counterparty name and I get back this list of like dozens and dozens of documents. I'm like, oh my God, what are the terms? What are the actual operative terms? And there's no document that gives that to me, shockingly. And so that's what Contract Compass does. It takes the various pieces of a contract and understands using AI what has been superseded and is able to like give a change log and give a final version of the contract. It's um I didn't going into it, I wouldn't have branded it as GitHub for contracts because I didn't understand what GitHub does. But for people who know GitHub, which I am learning a little bit, it is GitHub for contracts.
Cecilia Ziniti:Wow. So you created this, you coded this yourself, and it's like a landing page. I'm in legal at Instacart. I go to this, I pull up, you know, let's say you got your deal with, I don't make this up, Albertons. You search for Albertons, you're gonna get every version, you're gonna get like a Git log, basically, of every version of that contract.
Nicole Altman:It is not integrated with our contract data at the moment, but yes, that's the idea. Right now you have to like upload documents to it, right? But yeah, you can upload the documents and then it comes back. You can see uh document by document change log, you get an overall summary of the changes, and then you can get a final stitched together version.
Cecilia Ziniti:Oh I love that. I love that. Yeah, so uh on the GCA, people have been asking for this type of functionality that is integrated. We've been calling it contracts nirvana, and I'm like, what do I call it? Because I think you're right that the state of CLMs today, unfortunately, is that they don't they don't do this.
Nicole Altman:Uh and so it seems kind of odd.
Cecilia Ziniti:It's shocking that it's 2025 and you know, uh a tool that you you know charges X amount of money to manage your contracts would not allow you this functionality. It's coming. You know, I could tell you from the AI side, we're working on it. We have very similar functionality where you upload, you know, this is a great AI use case where you upload to GCI or your tool of choice, like, hey, you know, here's our contract with acne. It's been amended 19 times with three spin-offs. We did have that with a customer where their primary agreement with this particular uh partner went back to 95 and had been amended six times, 26 contracts, three spin-offs. And the AI gave exactly what you described, Nicole, where it was like, okay, the operative contract is this 96 version, it was amended this, it was amended there. And, you know, to your point where like this used to be someone's full-time job, the person who managed that contract, that relationship, like the primary lawyer on that deal or set of deals, was like, this would took me like a week when I first joined to like marse the contracts. Here's contract A, contract B. Oh, wait, that was like actually amended and restated and this whole thing. This kind of I love that. This is such a good example. And I might have to steal the branding. So we'll look out for that. Just kidding, just kidding. We come up with our own thing.
Nicole Altman:But that but love that. So the conversation also, you know, brings up to me there there's so much expectation and hype around AI. Not that I think it's overhyped, but like sometimes like you expect it to be able to do so much. Like you asked, like, oh, so it's like in, you know, like if it's in your system. It's like, no, not yet, right? And because all of these things take so much time. But I think a lot of people have this expectation, like, well, I hear all of this hype, like I should be able to just do this. Yes, exactly. Like everything magically happens at once, and then when it doesn't do that, it's like, oh, this is disappointing. It's like, no, no, no.
Cecilia Ziniti:We're using fax machines. Yeah, no, I'm with you. And and to me, like, I actually kind of love that because it reflects back to these kind of leadership themes that you all talked about of growth mindset, where you know, they teach you in the book growth mindset, it's like, look, you know, it's not that you can't do it, it's that you can't do it yet, right? And so, like, get there. And yeah, I love it because to me, it's always like when attorneys ask me like the kind of question I asked, I just did it. I literally just asked you that kind of question. I'm like, soon, like, trust. It is still better. You know, it's kind of like not letting the perfect be the enemy of the good or the perfect be the enemy of the better, where it's like, okay, what systems can we put in place? This is so much fun. I am I am just like loving this. I'm getting like the time queue here.
Kelly Noguchi:But I do, I feel like I always want to like pick up like where I don't know, maybe it's like just the nature of my role. Like, I want to pick up where like both GCAI and like contract compass like kind of leaves off, right? Like you guys create these cool tools and then then what, right? Like I feel like I'm like, there's that friction point for me, right? Like then what? And so like to all those questions about actually operationalizing and adopting and stuff like that. And like all of the questions I think about when I'm like in like existing workflows, existing clients, like, and then like the broader landscape of tools available. Like, what does it mean? Like, I feel like with the transparency of each, even like choosing what model you want to have behind. Like, I feel like, oh, they're deprecating Tat GPT for like one of these for oh one of these. Like, I feel like there's all of these questions, like like to your point, like I feel like AI, so like I use cur like when I'm in cursor, I feel like, what the hell? Where was this my entire life? Like, why am I playing around with stupid dashboards like when I could just create one by asking this cursor to do it for me? And but like there's these like realistic issues like now what, right? Like, okay, well, I want to share this. Like, this is you know something that Nicole and I are like working on. Like, how do we internally share it like securely and then in like not losing anything, not like disrupting, like if we can't teach people how to download a repo. Like, like you can't, I feel like there's these like realistic issues that come up when you're trying to like scale something even to a smaller lot. Like, I don't I always like I can't like I feel like it would be remiss of me if I don't like mention this portion of it. It's like both makes it hard because like AI is like both like limitless. Yeah, yeah.
Cecilia Ziniti:I mean look I and this is this is this is genius. I I hear you say that of like what now? And that's legal ops' job, right? So we've had a bunch of conversations on this podcast around, you know, kind of adopting AI, around, you know, how do you empower legal teams to move fast and do, you know, meet the business where they are and in fact advance the business. And so when you're saying like, okay, the what now of like the realistic, I think that's where people, and maybe the MIT study is getting at this, it's like, okay, fun. Nicole and I are having fun, like you just can't contain us. We're like falling, we're falling off the screen here, so excited about the stuff we can do. But then it's like, okay, how do you get the long chain of the whole department of like the job has to get done? And so like translating that, I think it's for me, it's it's actually what you're bringing to mind for me is like when I as a CEO am like building a team, like, how do I have those super early adopters that move fast, you know, and kind of like want to do the new thing, people, and yet have the operational excellence to, you know, in your case, at Instacart, deliver, you know, hundreds of chargers to people that forget them at their hotel, you know, every day, right? So, like really getting that that business leadership. So, what what are your tips, you know, as we switch to slightly more personal, Kelly? What are your leadership tips? So you've had this great career at these like fabulous tech companies. What, you know, what would you tell your listeners are your general, maybe non-AI leadership tips?
Kelly Noguchi:I mean, I feel like the AI like an environment is like a really good, I feel like just like feeling like you want to learn constantly. And then just like living that each like challenge feels like, ooh, like new thing, like okay, what what's next? And then kind of bringing that because you know, like I I feel like a lot of times I'm what is it like there's a term for when you're persuading sideways, you know? Like uh like influence, yeah. Yeah, like yeah, like is that like horizontal influence, like uh leading by influence. Yeah, yeah, something. And I feel like a lot of that is like just like getting that energy alignment. Like if people are motivated to like work with you and like like get excited about that goal, I do think like that's you know, like everyone's so busy, like getting that like that meshing well together and then pushing it forward. So you know, do like really like feeling that for yourself about like trying to be engaged with like well, I feel like the world, but um just like all of the new technology or processes that are out there, and then trying to like get other folks like in that same like space as you so you're not just stuck on your own.
Nicole Altman:So yeah, I love that. Just like being being open-minded and having a curiosity and just a learner's mindset, I think is so important like with everything, but especially with AI, because it is all so new and you know, there's there's people calling them themselves experts, and there certainly are like AI researchers who are are are true experts. But I think most people like nobody's really an expert yet. Leaning in, learning, experimenting, but being curious and open-minded and willing to like get hands-on and explore, I think is um is what I don't want to say makes makes you an expert, but I guess it's really the only way to do it.
Cecilia Ziniti:How do you get there? Like the the kind of this journey aspect. No, I love that. So on that, on that note, Nicole, what's a book, podcast, article, something where people can go? Maybe it's the MIT study, but what it what is somewhere where people can go and think about how, you know, something that shaped you and any book that shaped you, either on the tech side, leadership side, legal side.
Nicole Altman:I really enjoyed The Worlds I See by Fei Fei Lee. It it blended a lot of the human element to um also just the we talked about already, like the grit and perse perseverance and determination that it takes to push the boundaries. Um and so I I really love that book. I I mentioned earlier I love podcasts too. I I I listen to all manner of podcasts, but I particularly love listening to ones that are not necessarily tech focused or lawyer focused, and just kind of get that broader perspective. And, you know, Radio Lab is a podcast that I've been listening to for a long time, and it's just filled with like, you know, investigative journalism, taking these big ideas, often science or technical, but presenting them in a very human and real way. They do have a legal spin-off, more perfect, which was exploring Supreme Court cases in a way that we don't learn about them in law school, really looking at the human element, the story behind the story. But I I think it's very important to have that broader perspective and then be able to connect the dots with like something you you learn here in this other domain and how it might apply to my legal work, my explorations with AI. So I love that.
Cecilia Ziniti:Kelly, any any, any, any podcasts uh that you have in common or any anything to recommend on top of the the the great ones that Nicole that Nicole shared?
Kelly Noguchi:Yeah, like same. Like I feel like I definitely love to listen to podcasts and shout out to definitely anything NPR. But um, and also like but these days I feel like I'm a little bit hooked on like TikTok and stuff. Like I then that's where I'm getting a lot of my like tips or like new like because it's changes so fast. I feel like even podcasts that come out like every week or you know, even daily like sometimes aren't keeping up. Like I heard the term MCP like months ago, and I'm like now it's kind of blowing up and like kind of more main more mainstream stuff, but like I'm like what it you know, and people are just like, oh that what MCP, MCP, like I'm like, what is that? And that you know, on TikTok, just users are like talking about how it's been incorporated and into their ecosystem, and like and so and you know, obviously you have to be really careful about um what you like learn on TikTok, but like I feel like I it's just something like really easy to be like that seems really like on point and like like live basically. And so you know I've been kind of hooked on there's like a few like AI folks that are just like normal folks that are you know like like whatever data analysts in this place or just AI enthusiasts, you know, that are very interesting uh creators for sure.
Cecilia Ziniti:I love that. So this podcast, y'all are so kind of like tech forward and love to learn just making me so happy. All right. So closing reflections, uh Nicole, if someone listened to only this conversation from you, um of course they can follow you on LinkedIn, they can find you, but anything that that any nugget around them understanding using AI in legal or in general how you approach your job um as uh senior counsel at Instacart?
Nicole Altman:I think just being willing and interested to get hands-on, deeply understand the business, deeply understand the systems, deeply understand the tools. I apply that to my work generally, to my use of AI. I love to just get in and figure it out.
Cecilia Ziniti:I love that. Kelly, how about you? Any principle or mindset you want folks to take away listening to this?
Kelly Noguchi:Uh I mean that's I gotta say, like really, like learning, like the learning mindset and like part of the challenge, I think, like right now for me in particular, I guess, like is like how do I like add in that learning mindset to like the organization? So like, you know, we're learning fast enough and adapting fast enough. And like to like so it's not really about the AI, it's like just us like just you know being ready for any changes because it's gonna come.
Cecilia Ziniti:Love that. Change is gonna come. All right. So the two of you have shown us how legal teams can move beyond AI curiosity into real practical impact, right? How leadership comes from building with both ambition and empathy. Where can people connect with you to learn more?
Nicole Altman:Reach out. I'm on LinkedIn. Happy to connect.
Cecilia Ziniti:I can assure you that you all are quite a bit ahead. And with our listeners, we're kind of all on this journey together. So thank you all for taking this journey with me. That was my conversation with Nicole Altman and Kelly Noguchi. They're two leaders at Instacart Legal. We're proving that AI in Legal, it's not hype, it's real tools, it's contract compass, it's moving forward, and it's systems making an impact and a difference both in AI adoption and across the business. So if you're thinking of bringing AI into your own practice, whether through small experiments, cultural change, fun, learning mindset, or building whole new systems, I hope this episode gave you something to take back to your team. Follow CZ and Friends wherever you get your podcasts. You can subscribe to our newsletter at gc.ai slash newsletter for more conversations, behind the scenes tools, and takeaways for you and legal leaders in general. Thank you so much for listening. Have a great day.