DraftAId, by Mohammed Al-arnawoot

We explore how DraftAId automates mechanical drawings from 3D models, why focus beats hype in CAD AI, and how human-in-the-loop design keeps engineers in control. A live demo shows associative drawings in Inventor, and we preview cost estimation and an upcoming open version.

• Drafting automation for mechanical fabrication
• Lessons from YC to 175k+ drawings generated
• AI hype versus real manufacturing workflows
• Human-in-the-loop interactivity and PMI implementation
• Datum strategy, tolerances, and communication intent
• Integrations with major CAD and vault systems
• Competitive landscape and why focus matters
• Preview of cost estimation for North America
• Shift from enterprise-only to open access

Transcript

Welcome And Guest Intro

Roopinder 0:00

But the so the heat. Hello, and welcome to FODES, the Future of Design and Engineering Software podcast. My name is Roopinder Tara. On the show, we will have guests that will discuss tools and technology that engineers will find interesting and useful. I'm delighted to have as a guest Mohammed Al-arnawoot, co-founder of DraftAId. I met Mohammed over a year ago, and at that time, DraftAId was still in Y Combinator, the Silicon Valley's most famous incubator. Just admission to the YC is a thrill for any startup. Since then, DraftAId has graduated from YC and moved back to the Toronto area. If DraftAId makes you think of a benefit concert for unemployed drafters, you're sort of on the right track. DraftAId is bent on making the drafting of 2D views the bane of an engineer's existence almost completely automatic. Mohammed has a bachelor's in mechatronics from the University of Waterloo in Ontario, Canada. He has interned at Nvidia, been a VP in another AI startup, and served as a director of platform engineering for Dozer, a search engine for You guessed it, heavy equipment. You joined Mohammed in his Toronto office.

Mohammed 1:40

You were one of the first people to write a feature article about us. We were at Y Combinator. I think you found us, I don't remember how, and you wrote an article about us in engineering.com. Where are you based on? 

Roopinder 1:51

Near San Francisco, north of San Francisco. How about you?

Mohammed 1:55

I'm in Toronto.

Roopinder 1:56

Mohammed, do you want to take a moment to just introduce yourself to the podcast audience? Yeah, absolutely.

Mohammed 2:02

Mohammed, I'm a mechatronics engineer from University of Waterloo. I'm the CEO and co-founder of DraftAId, the world's first AI to automate engineering drawings, mechanical fabrication drawings. You know, DraftAId is automating that tedious process of going from 3D to 2D and engineering engineers design into anything from skyscrapers to automotive parts have used our product to generate hundreds of thousands of drawings to date.

Origin Story Of DraftAid

Roopinder 2:27

This first podcast we've done with you, that's for sure. Podcasts are hot right now, so we're gonna have a little fun. Are you able to show us a demo of draft date? We can do a demo date. Tell me a little bit about how you got started and how draft date came into being, what his original purpose was, and the most important question of the day how do you use AI? Everybody, no conversation is complete without talking about AI.

Mohammed 2:52

We put AI right in our name, DraftAId, and there's an AI right there. We're definitely in a bit of a bubble right now, and we can talk about that. We started DraftAId about two years ago, actually, I think our two-year anniversaries in like seven days or something. We started two years ago, and it happened when I was driving back from a camping trip, and a buddy of mine told me, Hey, Mohammed, you gotta meet this guy. I met what is now my co-founder, Abdullah, who is a drafter at Harman, and him and I were chatting, and he said, dude, I sit in a room with 40 drafters, and our day-to-day is to take these 3D models and make drawings for them. And it's just crazy to me that this is what we do. You know, you zoom into a pixel, you click, you zoom out, you go to another pixel, you click, and that's all I'm doing. And why is this not automated? You're telling me Chat GPT can write my emails and presentations, and I gotta zoom in and out here on pixels. And so I said, Yeah, that is a pretty interesting problem. In a previous lifetime, I worked at another startup that I was a VP of end where they ended up getting acquired ultimately by Autodesk. I'm a startup person through and through. I've worked at a couple of startups in my career, and so we said, let's figure this out. We had our third co-founder, who we have the most Canadian founding story where early on we would draw algorithms, our ideas in snow on hikes. It's probably the most Canadian story there is.

Roopinder 4:06

We're got a hockey skate, probably, correct?

Mohammed 4:09

We thought, why not just go back to you know, even before paper, you draw right in right in the snow. And so we tinkered with the technology for a while test in our team has a PhD in AI from University of Toronto. We looked through the technology and started the company officially in October 2023. A month later, we got into Y Combinator. So we did Y Combinator in San Francisco in winter 2024.

Roopinder 4:29

You're having an anniversary right now.

Mohammed 4:31

Like it's in seven days. That's fun. We did Y Combinator and launched with a few customers back then. Since then, we've been on this journey of automating drawings. We raised around from industry, folks who've actually worked in this industry in the manufacturing space. Our backers, our partners are actually in this space. In our first year, I remember we auto-generated about 10 to 11,000 drawings across all of our customers. I thought, well, that's amazing. We generated a lot of drawings. In our second year now, as of two days ago, we've auto-generated 175,000 drawings across all of our customers.

Roopinder 5:07

Think of all the pain that you've alleviated with that, right? No, I don't think any engineer likes to make drawings. And I don't even think people that are paid to do it, like mechanical does like to make drawings. Now everybody wants to work in 3D, right?

Mohammed 5:20

Well, yeah. I just jumped off of a sales call with a customer. And I think one of them, I think they're based out of Texas, and they said something really cool. He said, Yeah, when I'm done with the 3D model, that's when the fun ends.

Roopinder 5:30

So now, how in that time, by the way, congratulations, that's an awful lot of drawings. And it seems like you've gotten pretty successful since the last time we talked. Are you still dependent on your initial funding or are you generating enough revenue to proceed as a business? You might still be in the growth phase.

Mohammed 5:46

Well, the reality is that when it comes to startups, especially in our time a lot of hyper AI, but the thing we're seeing is we we've 10x'd you know, our drawings more than 10x since last year. At the pace we're currently going, the demand is outpacing the supply, so to speak. It's a great problem to have for us.

Roopinder 6:02

I imagine a lot of people that maybe didn't even know how to spell CAD before are now like, whoa, you are using AI. I think we ought to take a look at this. That's how hype overheated this AI market seems to be.

Mohammed 6:14

It's 100% a bubble. I see an AI company in the CAD space pop up like every other week at this point.

Roopinder 6:20

Yeah.

Mohammed 6:21

And it's interesting because if you actually look into, I think there's a lot of folks who are coming from a software background into this industry. Right. What works in software in terms of AI tools does not translate into that mechanical engineering industry or manufacturing. One of the blessings is having someone like Abdullah, my co-founder, because you have to have industry experts, people who put in their 10,000 hours, so to speak. Abdullah's usual joke is I got carpal tunnel making drawings. You need those folks in the room. And you obviously need folks from software engineering background to build grade products in here, but it goes back to workflow. If you're building a tool that doesn't fit into the workflow of a mechanical engineer or a manufacturer, I usually look for companies that are doing something real. They're laser focused on one problem in the supply chain. It might be sourcing parts, it might be 2D to 3D or 3D to 2D. Focus on one area that shows me you know what you're doing, and you know how hard it is.

AI Hype Versus Real Manufacturing Workflows

Roopinder 7:14

It's a hard problem. I hated my drawing, but I had to do it. You had to cut your teeth on making drawings first, and then you got onto the cool stuff, the stuff that was fun. It was really hard. I hated it. I wanted somebody else to do it for me as quickly as I could. Nowadays, you're right. There's this influx of companies willing to take this on, and I'm so happy to see it. Do you find that they've competed with DraftAId now because they have discovered that they can also make drawings?

Mohammed 7:41

I've seen it up since we started. When we started, it was just around the time when the only other, I think, drawing related, serious drawing-related solutions were maybe DriveWorks, which most folks are familiar with. And yeah, that's the it was around the time when Fusion announced their drawing automation tool, the one in Fusion 360.

Roopinder 8:01

Right.

Mohammed 8:01

That was it, you know. And DriveWorks is not an AI-based tool, it's usually a highly customized tool, very particular things.

Roopinder 8:07

So rules-based applications, it's very difficult. It's a lot of programming, it's a whole expensive big application you have to buy. But I've heard that from Autodesk primarily, that Fusion is going to give you a run for your money. They've got 2D drafting pretty well along. And that's a major company that's now stepped into the ring probably like a year or more ago. But just the other day, I heard I was talking to guys at Solid Edge. They're making a big deal of how they're automating the drawing procedure, right? And all the time I question it and I want to proof as they say, it is difficult to make a 2D drawing. The people that are really good at it consider it an art to get just the dimensions exactly in the right view and not over-dimension. Certain things I guess are easily automated, like penmanship, right? Lettering, right? But otherwise, there's a lot of thought that goes into making a really good 2D drawing. Yeah, so I imagine with your professional experience in the field, you must have brought some of that into the product, right?

Mohammed 9:06

I think there's two elements to it. One is what industries are you building for? Building for aerospace is not the same as building for automotive and construction or woodworking. These are all different industries and they have their own nuances. That's why trying to build something all-encompassing doesn't work. It's pretty hard with the current technology. So focus is important. The second part is being a mechanical engineer. It's not just lines and isometric views and projections, it's a tool for communication. And you're communicating often with an NMA Seattle, and you're sending this drum to someone in China. And that's that email with that PDF is the only communication you're gonna have about. There's a lot that goes into this, what you're showing, what you're not showing. Even things like how many decimal points are you putting on that dimension? Well, that matters.

Roopinder 9:50

Yeah, it does. I know. I've gotten thrown out of the shop for using too many places after decimal point. But even other things like, okay, I had a lot to learn. After I thought I learned everything in school, one of the things I had to learn was, like I said, no, don't over dimension. Don't put too many decomals, don't put too high of a tolerance on it. And know which dimensions are important. Like you don't want to, you may not want a dimension for mechanical drawing always from the corner of the part. It might be the first hole that's the most important, right? And then you want to index everything off of that first hole, right? Because that sets the plate or bracket or everything in its place. The corner may not even be that important. How is the AI gonna know that?

Mohammed 10:30

I think AI tools in general have to have. So, for example, let's take the datum example. It's a great example. I think AI tools have to have two modes in which they operate. One is, hey, I think this is what you should do. I think this should be your datum. I think this should be, and it's predicting it. It's just guessing, right? And all AI is just a really good guess. Right. But then it also has to have a mode where it's like, I don't like your guess. I'm gonna change it. Give me that data.

Roopinder 10:54

Why can't it interact? Be more interactive. Why does it have to be a one-shot deal where it probably is gonna get it wrong? Why can't it just say, oh, wait, is this where you want to start? Is this your datum whole? Is it yeah? Do you do that?

Mohammed 11:05

Because if we do that, so right that is it. After what is it, 185,000 drawings or whatever, 186? We've talked to so many customers, and there are companies with all say, you know what, our parts are simple, throw it into the AI, and we're happy. And then there are folks who are like, I want this data for this part, I want that data for this other part, I want this kind of tolerance stack, and I want to give you input, and then you save me the thousand clicks. Right. And so we do both, and you just have to this is where the bubble of AI will replace engineers. If you actually see most AI tools in action, some of them are pretty funny in terms of results. Right.

Competitive Landscape And Focus

Roopinder 11:38

Oh, yeah, yeah. There's some really goofy results you could get. I've seen quite a few mistakes, and unfortunately, that tends to turn engineers off of AI in general. This is what I'm what I'm trying to do is like, this is what works, this is what you need to look at. Whereas they're getting a barrage of information about AI, and most of it is not put out by engineers or or people in the field, right? No industry experience. So they're I would say this even about software companies. Software companies making CAD drawings or products are not necessarily involved in the industry. There are companies that don't have engineers at the top or mid-level or even as product managers, and they're software people, right? So they don't know the market. And I always think, oh, I don't know if you're gonna get that right. I always have this skepticism. A mechatronics engineer, I learned mechanical engineering and software.

Mohammed 12:27

I happen to be one of the few people who's in both seats. But we said, okay, well, is it easier to find a mechanical engineer and teach them software? Find a software engineer and teach them mechanical engineer. Yeah. And this now is an interesting question. The answer we found is one, building software in this industry is really hard. If anyone here has worked with Parasolid, anyone here has coded on any kernel at any point, they will know how difficult this market is. When I see like really big claims from big companies, I'll say, I don't know, show it to me, seeing and believing gear. Then our solution to this problem was the concept of forward deployed engineers. Okay. And so what we do is our software engineers go and sit on customer sites, they interrupt the mechanical engineers, they see their workflow, they go through drawings. We have this meeting that we do with all of our customers, especially the large customers, where we go through a drawing review with them, and it's this like three to four hour meeting of going through drawings together. You need a lot of coffee for that because it's not the most exciting thing.

Roopinder 13:24

We've gone past a few people's attention span by then.

Mohammed 13:27

Oh, 100%. I think I've seen so many drawings at this point, and our software engineers who have no mechanical engineering background have seen so many drawings, and that's how we make sure everyone across the team, you've seen a machine shop. If you haven't been to them, you need to see a machine shop. You need to know what a CNC does, you need to know what a late does, you need to know why they're used. It's an interesting job.

Roopinder 13:45

How do you feel about a company like Autodesk overpowering not just you, but a lot of these startups that have really good ideas and may have better implemented those ideas, Autodesk, or they being DASA systems, or they being semens, okay, have made an okay attempt at it, but now are going to be seen as known quantities.

Mohammed 14:04

I think it's a matter of relationship with a lot of these folks.

Roopinder 14:08

Okay.

Mohammed 14:08

We have a good relationship with the folks at Autodesk. We chat with them frequently. In fact, I think our first hire was someone from Autodesk. We have a great relationship with these guys, and I think it's important to have that in the industry.

Roopinder 14:19

Do they have a third-party partnership? They do. Okay. Are you part of their partnership?

Mohammed 14:24

So our Autodesk is actually, I'd say, probably one of the most forward when it comes to partnerships. And it's really easy for us because they have an office in Toronto. We're in downtown Toronto. But I'd say Onshape, Creo, PTC, so Siemens. I mean, I think when we started drafting, around the time when you did the first interview with me, within a week of that, I met with the head of AI and innovation at Siemens at NX.

Roopinder 14:46

Was that coincidence, or do you think we had anything to do with that?

Mohammed 14:49

I think your article definitely brought eyes to what we were doing. The other aspect of it is you know, Fusion, Revit, Inventor. Sometimes those folks are thinking architecture, construction, right?

Roopinder 14:59

Very different. Very different. Every industry is different, woodworking versus ordinary mechanical design, let's say. Architecture is like a whole different beast, right?

Mohammed 15:07

It it really is. We don't do much in architecture. We work with some companies in constructions that are doing more typical mechanical engineer, but architecture is one world.

Roopinder 15:17

We know our focus, we stay in our lane. All right, good. That's wise. Until like second round of funding, perhaps. Because your name doesn't bind you to just mechanical, right? So no, it doesn't.

Mohammed 15:28

We get a lot of our customers, and a lot of the meetings we do are doing maybe 50 sales meetings a week, and a lot of them come from architectural. And I tell them, I'm really sorry. We just can't help you. And they'll say, But it might look like it, but behind the scenes it's not.

Roopinder 15:41

How hard could this be? Just make it feet and inches. Yeah, that's the boy basically.

Mohammed 15:47

I think my favorite line I tell folks is guys, it's AI, it's not magic. There's limits.

Roopinder 15:52

I would let you do a demo so you can show you. Yeah, happy.

Mohammed 15:57

Let's do this. We integrate with most of the big CAD software you're probably familiar with. Over here, I'm gonna demo with Inventor. I think I'm running in version 2025. Can you see my screen? You see DraftAId?

Roopinder 16:09

Yes, I see all the drawings.

Mohammed 16:11

Fantastic. This is the DraftAId UI. It talks to your cat system, in this case, obviously, in Inventor. And making drawings, we we our vision. And when we first started the company, we said you want to make ideas and CNC machines closer. And you want to just bridge that gap. And so drawings was one of the first things we started with. And and we've expanded our focus since you and I last talked. I'll show you what that looks like. But the idea is you can tell the AI I want these drawings, and that's what we'll do right now. You can also have it run behind the scenes and watch in Inventor model, change, work on your modeling, and then you can generate drawings without you even telling it. Without further ado, let's jump into this. So this is the draft AD UI. These are a whole bunch of drawings that I had already made, but making more drawings is as simple as just grabbing and dropping it here. It's a little a bit of time here. Typically, a single drawing takes anywhere from 30 seconds to maybe a minute and a half, maybe two minutes. This is on my template, but obviously when we work with customers, it looked like any of their existing drawings. You can do part drawings with DraftAId, and we'll talk about some of the other drawings. You can also draft it.

Roopinder 17:17

Is it being done locally or is it all cloud-based?

What Makes A Good Drawing

Mohammed 17:20

Aspects of it are cloud-based and aspects of it are local. I'd say the heavy lift is on the cloud, and then the general drawing generation, because we need to conform to whatever your CAD software does. I booked the number of drawings we generated, and about 83% of them were all on the cloud. We do have an on-prem solution, and some of the big guys, like automotive, won't be able to do that on on-prem with some of those folks, and that's fine. So we support on-prem, but I'm surprised at how many folks are fine with cloud. We make sure that you know we're GDPR compliant, we're SOC2 compliant, and everything in between. So, anyways, we have our first drawing. It's took about 29 seconds to get generated. Quite a few holes here, and so it decided to add a whole table. Whoever made this model did make good use of whole wizard, and they didn't just cut and extrude. And so we get the benefits of that, obviously. Everything on the drawing, from sheet size to which views to show, which views not to show, what to dimension, what not to dimension, all of that is decided by the AI, right?

Roopinder 18:18

It said that it made use of the whole wizard. So I would look for if I was seeing how automatic this is, that the hole would would be a hole call-out rather than like giving a diameter and a depth separately. But is that I don't see that to be the case. Well, it's right here, it's in the whole table. Right in front of me. So if for because I'm seeing the section view that shows it looks like a hole with a diameter in it.

Mohammed 18:42

To show this hole and this feature, we got a little bit of a jagged section view here to show these two features. Right. Usually this particular AI, whenever we work with a customer, as they're using it, it's seeing the changes they make to their drawing and it's learning based on that difference. So I'll give you an example here. This is an actual inventor drawing. I can move it around, I can change it. This is an Inventor drawing. This is where I tell people this is a human in the loop problem. Because after I generate the drawing, we always say there are two engineers sitting next to each other in a company for 20 years, and they'll disagree on what's a good drawing and what's a not subscribing. No matter what we do, at the end of the day, the engineer needs to be able to go in and make changes to this, make any edits they'd like.

Roopinder 19:23

And I could go in there and move dimensions around, they'd still be associative.

Mohammed 19:26

Absolutely. This is associative because we're using your inventor. It's associative to the actual 3D model. It can do it within Vault, so connect to Vault or PDM.

Roopinder 19:35

Was I always in the Autodesk Inventor?

Mohammed 19:37

No, I took you there. See there. If we click this button here, let's go to another one. Let's go to the I think this is the same part, just it's it's mirror. But if you click open, this opens it up in in Inventor for you.

Roopinder 19:49

Okay. So it does its magic with the DraftAId aid, and then the drawing is fully available. It's still an inventor drawing, and I could still access it in inventor.

Mohammed 19:58

Absolutely. So this is mainly used to orchestrate the dance, so to speak. And it's what talks to our servers. But then at the end of the day, the file is on your computer or in your vault, and it's your drawing. You can do with it whatever you want.

Roopinder 20:11

An IPT, that's an Inventor file. Yeah, that's an inventor file. If I click opener, we're back in Inventor. I'd be hard pressed to find a fault. I mean, I probably could if I looked at it long enough, but still.

Mohammed 20:21

There are faults. I'm I'm aware what the faults are, but I'm not going to call them out. But there is a salesman. I mean, realistically, this is uh it is an AI product, so it is a probabilistic result. We've been running this for two years now, and it's just getting better and better. We're lined up for a pretty big launch in the new year. What we've learned is mechanical engineers have a very low tolerance for mistakes because our work is hard. You need to be pretty darn close to correct. And so we get you about 90% of the way there, and that's our promise. But there's always a 10% that at the end of the day, you're the engineer, you're the expert, right? You know the part better than I ever can.

Roopinder 20:58

And so you shouldn't make those last touches. And I have a history with that part or similar part. So I kind of know what I want. That has to come across. Were you at Autodesk, the last Autodesk University, or it's being called AU now?

Mohammed 21:10

We were not, actually. I was in Europe that week.

Roopinder 21:12

So what they told us there was, and they've said this before, is that automatic drawing is only 70 to 80 percent good. Now I know these are approximate numbers, but you being at 90, seems like you've passed them. They totally concede that there's more work to be done when their AI is done, which is also what you said. Have you compared the two? We have compared the two.

Mohammed 21:33

I think we get away with a hack that a company like Autodesk doesn't, which is anyone under the sun can use fusion for whatever reason. When they built their AI product, it needs to work for the architect, for the woodworking guy, for the guy making doorknobs, just as well as the guy making a drive shaft, right? We are more selective, we don't work with architecture, doorknobs and those kind of things. Pretty selective in what we do. And because we have more focus, we can get higher results.

Human-In-The-Loop And Interactivity

Roopinder 21:59

If we widen our focus, our results would also drop to probably the same. So you can extract that extra, I'll say roughly 10% better, more completeness.

Mohammed 22:09

I genuinely do think I've spent a lot of time with fusion. I do think the Fusion team did probably one of the best jobs with an AI tool right there on the reverie AI tools you can just go and use. Yeah. And there is a hey, we're showing you our work. It's right here.

Roopinder 22:24

I wouldn't encourage you to look at Solid Edge too, because they have a history of doing drawings really well. They can go up against Autodesk for their drawings, and now that they can automate the process, I have to look at it too. It could be pretty good.

Mohammed 22:39

I just had this conversation yesterday, so too early for me to actually test it, but there isn't one one other feature I definitely would like to show because this is new, this is still cooking, it's not out yet, but it's something that we're working with a few folks on, which is what we call auto estimation. A lot of reasons why people want a drawing is sometimes to estimate the cost of the part. And we've been working on our own model for this where you can essentially tell me what is the quantity of this part, whether you need it urgently or non-urgently, and estimate costs. Oh, and it will give you the cost for producing this part. Now, this is only going to be available in North America to start. It won't be available outside Canada and the States, but it's something we've been working on for some time, and it's probably one of the highest asked for features.

Roopinder 23:22

Is it accounting for the materials as well as the machining operations, the time it takes, that sort of thing? Absolutely.

Mohammed 23:29

Material is captured in machining operations. What are you drilling? Is this a 3-axis? Is this uh indent and cut part? Is it going to go on a laser? Urgency factors in. Are you getting this in a couple of weeks or are you getting this in over a month or you know, two month timeline? And that for most folks, that changes the price point.

Roopinder 23:45

Wouldn't it want to know or ask if this is going to be machine? A casting would change everything dramatically. Over here, we show you kind of what is detected once again, right in front of me.

Mohammed 23:56

We haven't made this very interactive just yet because the underlying technology isn't actually interactive. As we roll this out, we will see what the industry thinks. Again, this is something where partnered with a few companies to be the first people to try it. We'll get their feedback and then iterate on it. But I'm pretty excited. This is one of those things that is going to come in the new year.

Roopinder 24:13

You have to excuse me for asking about interactivity. I know I understand it's difficult. I was happy to see when uh Chat GPT came out. If you ask it a couple of times, it starts thinking more and actually asks you, wait, do you want this or that? Then it starts interacting. But I can't turn on the interactive all the time. Most of the time it assumes I want something. And you can say, right, but I wouldn't want to use something like that on a real engineering application. I'd want it to be more of an assistant, right? More of an aid, AIDE. Or like, okay, it's guessing that this is the primary view, this is the datum, this is the most important whole, that sort of thing. That's what I want. Zero experiences seem to have infused into the product, helping quite a bit. But still in my mind, if I was shopping or creating, I'd say, let's make it interactive too.

Mohammed 24:59

Um interactivity. The way we did interactivity is actually based on there's tools out there where I've seen folks where they take you out of the cat system. Yeah. That's where you do the interactivity. We didn't like that idea. We found that folks didn't really have much of an appetite for it. So what we actually do is we have a few different tools. One of them is you can go into Inventor and color a face. Or for example, in SolidWorks, you can datum a face. And if you do those things, that's how you can say, hey, you know what? This is what I want. Use this color for a certain tolerance, or use that color for a datum face.

Roopinder 25:34

I see.

Mohammed 25:35

Or you can do just your usual PMI and BD tags that are within, for example, so can we actually read those?

Roopinder 25:41

Oh, okay. Factor it in. That's good. Okay, yeah. So you actually can read more than the geometry. You read the metadata. Okay, cool. Absolutely.

Mohammed 25:49

We're we're we're fairly integrated into it because the engineer is already working in that space, right?

Roopinder 25:54

Yeah. Why waste that information? Right. No, that's great. Okay, that's where the industry experience and software experience comes in handy. Exactly. That's great. Is there anything else to that costing thing? Although I think that's a big job.

Mohammed 26:06

Yes, in terms of the big news and what we're launching in the new year, draft aid so far has been an enterprise-only product. Usually you have to go through a sales process to use it. But in the new year, we'll be launching an open version. Anyone can come to www.draftaid.io. They can throw in their 3D models and get drawings. So it'll be open to the public. Anyone can use it. There will be a pricing page, and maybe we can get you a free license to try it out.

Roopinder 26:31

Oh, that'd be great. I would appreciate that. And it sounds like we ought to reconvene that and see what you guys are up to. It seems like you're moving pretty quickly. We are moving pretty quickly.

Mohammed 26:40

I think last year our team was four, five people. And this year I think we're gonna end at about 15 people. Okay. Still staying in that Waterloo or Toronto area? We're in Toronto, so we're in downtown Toronto. Our office, if you're ever in Toronto, come by.

Partnerships With Major CAD Platforms

Roopinder 26:54

Well, yeah. Well, it used to be more often the case. Engineering.com's headquarters are in Mississauga. Yeah. Nowadays we can meet when you come down to Bayer. Yes, I'll be there in a couple of weeks. So I'll start with it. Drop me a line. Great talking to you. Thanks for the demo and best of luck. I think this is a very valuable tool that you're providing. This is what AI ought to have done, right? That if people were anticipating that it can add value, it can strip away things I don't want to do and make me more productive as an engineer. Thank you for that. Best of luck to you and to draft day as a success. Thank you. All right. Talk to you later. Bye-bye.

Roopinder 27:28

Thank you for listening FoDES, the Future of Design and Engineering Software Show, brought to you by ENGtechnica. I hope you have learned of a new application or technology that will help you with your job. If you have an application you think would be of interest to other engineers, please let me know by emailing me at roopinder@engtechnica.com or message me on LinkedIn.