iCIMS Hacks: API Refresh, Token Limits & Ghost Jobs (6/26/26)

System admins tackle evergreen job refreshes via API automation, unpack iCIMS's upcoming token enforcement policy, and dig into ghost job legislation hitting New York and beyond. Plus: CXM talent community tips and email deliverability best practices.

https://systemadmininsights.com/

Transcript

 Great, all right. Welcome everybody to system admin insights. Hope you're doing great this week. We like to kick off every meeting with a little bit of gratitude.
 What is one thing that you are grateful for today. And today, I am grateful for MCP. We have been having a lot of fun with Claude's MCP.
 Uh-oh. T-Frozen for everybody else? Yep. Oh boy. So what he was just talking about is we've been having a lot of fun with Cloud MCP in our organization, which if you're unfamiliar with what that is, it's multi-context protocol, and it's specifically invented by Anthropic, but it's the way a lot of different
 large language models are going to start working with different AI technologies. So just to give you an idea of what we're doing with it, I personally created a task called an orchestration task, where I ask it in the morning to on a simple prompt, and this was given to me, this idea was given to me
 by Caitlin, so I will not take credit for it because she did it first. But where I give it a simple good morning and it automatically goes out and looks at my email, my Slack, every program that I'm interacting with and calls out the different things that I have to do that day.
 So, so-and-so emailed you, here's the summary, go, here's a response that you need to put together, review it. It's just every thing, it takes that mental load of all the different places that you have to go to look to see all the different fires you have to put out and put it in one place.
 It's amazing. So Alex, I carried on with the MCP description because you pressed. Thank Thank you so much. Yeah, what a day.
 My PC is still not working. All right. Anyway, I was just saying I was grateful for MCP. We talked about MCP Vivian.
 Yeah, but I was explaining the orchestration skill I created after Caitlin showed us hers on Tuesday. Amazing. Nice. And so we've got, it's set up to where we've got a database in notion.
 As our core source of truth. And Claude's MCP can reach out to all of our different systems. And then funnel everything into notion to preserve IRD's institutional memory, which is something we've talked about institutional memory a lot.
 But, you know, now we can just type into Claw, go up, take Notion, such and such, right, and it does it.
 So I feel like we're just getting started with this, and the whole team is doing amazing things with it. So, very grateful also just to work with folks who are so into technology, not just from a technical perspective, but also from a human perspective and really caring just as much about the impact
 on how we work with each other and how we work with our clients as how cool the technology is itself.
 All right, Vivian, if you would share, please. So we're going to lie and call questions right off the bat. Well, actually before, scroll down a little bit to the recap for June.
 So share your June 2026 wins. If you scroll down just a little bit, let's see what we got. So, Michelle, Michelle Branchwag said I finally got some formula fields set up to automatically pull salary information and lead in hashtags into our job description.
 We standardize the JD format so I could segment the boilerplate language into different sections to make them easy to bulk update.
 Love that. We also adopted a retention policy and did the first large scale purge of candidates in the system over four years old.
 I'm sure Vivian is glad to hear that. This one was manual, so the next step will be to implement consent fields and create automation that uses those fields.
 Really, really cool. Thank you for sharing that, Michelle. Tony Fairchild. So Tony is on the call. She said launching the new interview scheduling tool with the calendar integration after working for two years to get everything approved and configured appropriately, shout out to Kate and fail for help
 with the final piece. It's so glad to hear that. So, Tonya, what took two years? So for us, all of our IT approvals have to go to Tokyo and they have to review everything you've got to put together a work plan and then they have to review it and suggest changes and it's just a lot of layers of approval
 . We also kind of hit or it's a silly roadblock, but the profiles we were testing with we couldn't get the calendar to configure it properly and it was because they weren't in a hiring manager person folder.
 So it took it took a few months and a lot of conversations to figure that out and we kind of stumbled across it.
 But we figured it out, and so we're going live, so very excited. That's fantastic. I saw a Japanese stand-up comedian a few weeks ago who was talking about bureaucracy and digitalization of paper processes.
 Believe it or not, this makes it into stand-up repertoire. And they said that in Japan, and they were, they rolled out QR codes as particular government agency.
 QR codes to do something online and you scan the QR code and it gives you a download so you can download the PDF and print it out and then mail it in.
 Yeah, yeah, yeah. So I mean, really it's, you know, there's, there's like the friction of bureaucracy really differs from country to country and that user experience.
 You know, we have a particular set of values where I think by default, we assume everything should be easy, you know, but that there's a it's culturally, culturally different.
 All right, scrolling down. Alyssa says we are in the final testing phase before rolling out contractor slash agency access to our ATS, which will allow HR and admins at the organizations we have contracts with to upload the letter of attestation for their employees during the onboarding process.
 This will eliminate the inconsistency and how we receive letters of attestation and our compliance team will no longer have to locate it and then upload it to items themselves.
 We're also going to give the agency's very limited access with the ability to view a few reports. Interesting. I've actually heard that to help eliminate back and forth emails with the help of dynamic filtering and access controls.
 See more, it looks like we will be able to manage one login group for all of our agencies and they will only be able to see what is relevant to their company, love it.
 We also got our hiring automation fix so we can start testing again to roll out the current version of the interview scheduling tool that will be tied to an automation with our start on start onboarding status.
 This will allow our new hires to self-schedule onboarding calls with our compliance team. This is a big win for our compliance team because it keeps everything in one system to help with coverage but also gives I'm a scheduling tool that they can receive support with.
 The current tool they are using has no support when they are experiencing issues. That is a lot of wins. So, right, Vivian, have you seen this agency reporting?
 No, well, I've seen agency reporting before where you can put in a report. It's just like any logging group, you can put a report on their dashboard.
 If you've got permission for them to see the fields and permission for them to see all the various pieces that are in the report itself.
 So that's not anything new. It's just not common. That particular login group, the agency login group is so locked down intentionally that most people don't put dashboard panels on that intentionally and honestly the more common thing I've seen with agencies is that dashboards are created without realizing
 that it's shared with agencies and they see stuff they shouldn't see. Yeah, so that was that was a constant problem whenever I'd log into a customer system.
 I'd be like, do you really want them to see this? Um, so Alyssa, I'm curious. Some of the hiring automation that you're mentioning, um, with.
 Well, I don't think she's on. I don't think she's on. I don't think on. Oh, yeah. Okay. I was gonna say, I'm curious how she's doing that.
 Some of the hiring automation, whether she's just talking about workflow status automation or she's got some kind of API the eye integration going on there.
 You're used to ask more. Yeah. Does anybody ever done what she's talking about? Mm-hmm. Yeah, that's really granular attention. Love it.
 If you have something you'd like to share with the group, feel free to raise your hand and share it today or post it on Circle.
 We'd love to hear about it. Okay, with that, maybe you can go out of there and click Live Call questions.
 All right. And let's see. I'm trying to figure out where we left off last time because I don't have mine in front of me.
 Maybe I don't seem to see the hearted ones. Why don't we start with some of the newer ones? Yeah. Well, I know Townsend has one and he's on the call, so let's start with Townsend.
 Hey, everyone, Townsend with L in here basically to summarize this long post, the TA team and I have been trying to learn how to use CXM in a way that could either replace or enhance the way we use the recruiter and indeed events was hoping to get some feedback or suggestions on just how.
 So you get new talent to sign up through this without just pulling them out of the platform already. Hi, you've done some CXM.
 Any thoughts on that? Yeah, I mean, that's definitely a challenge with CXM that I've experienced where, you know, who they give you out of your standard pool, right, they're pulling from the ATS.
 I mean, you can, you can bulk upload individuals, so all you need is an email and a name. It is set up, those campaigns are set up to be pretty flexible in that regard, and that's where I think you can see that win with kind of the funneling, especially using automated campaigns and engagement, that
 engagement tracking, so not just measuring, are they opening the email, but are they interacting with their website, are they interacting with their jobs, after we send them that initial contact, and then filtering them into different pipe lines, different email threads that are really encouraging them
 into call to actions. That's how I really think of using the campaigns. It's all about driving to that call to action.
 What are we doing with them? I know you get multiple talent pages, so that could be something that's a big win as opposed to just the one I know in like standard connect.
 It's pretty basic. You just kind of say, are you interested? These are job categories. Yes, no. And then it disappears into a talent pool.
 The design for CXM's talent management is a lot nicer. It's a lot cleaner. You can choose custom questions. You can make it as small as you want.
 You can make it as specific as you want. You can build those forms and drive them into pipelines and talent pools that are much more easier to access.
 Any ideas on like events like getting people to register for events that maybe aren't in the system already. So I know you can generate the QR code for registration, but any ideas on like kind of disseminating that or getting it out there for people to see.
 Yeah, I know the standard format is you have kind of a pre-registration link, and then you have a check-in link when you're at the event.
 So you do get two. So you could certainly send out that page. So as soon as you create the event, you get that registration page.
 So it's really built for like if you were running your own event, and you can send that link that web page out.
 you can brand it to that specific event. You can send emails to people inviting them to the event and then when they fill in that registration, that is a form that you can change around.
 Other than that, and then I mean you could upload people and then check them into events. So I know that's something I've run into before is where people don't do pre-registration.
 Like if you're going to career events, sometimes people aren't registering, they're just showing up the event and that's where you can kind of manually push them in and then check them into the event yourself.
 We've done a couple different things as well. Having the QR codes like posting a post on LinkedIn, if you're comfortable, you will allow it.
 Just asking them to join our talent network and pushing that out with the QR code. We don't really do the pre-check-in post-check-in, We just have kind of the one form that they fill out and then attach it to a pipeline to pull them in so we can follow up later.
 And we've also used a lot of our college recruiting functions when we have the QR code just sitting on the table to make sure that they ski and then capture and I think that's kind of what you're relating to are their tie.
 And then the other thing that we've done we partner mortgage industry is a little volatile. So, there's a lot of companies we partner with when they're having layoffs, and we give them registration forms to have people come into our pipelines and then kind of give them special treatment if they apply
 to or come into the registration form. That's awesome. Patrick, you had some thoughts on CXM? Okay, I'm going to try to talk, I'm going to listen to my voice, balance sciences.
 So what we did is we created talent community pages and we would put those on LinkedIn. We would also put those into email campaign blast and that is how we would draw new talent.
 But we would do a community blast and clear its jobs because a lot of our talent is clear talent and we were drawing in the new talent through that way also the same thing was sent into registration links to people that sign up for career affairs and stuff like that.
 Great. Thanks for sharing that. All right, Townsend. Any other thoughts? nothing here just it's been a challenge a fun one though I hear that all right well let's go back to June wins does anybody have a June win that they'd like to share with group something they managed to do an ice ends or ice ends
 integration with another system I wouldn't see it as a win but I'm trying to get some juice out of this lemon or make me lemonade out of this lemon.
 So today I had to quietly put a project on hold an Isom's project. Not because of Isom's. Isom's has been a great sport trying to get it.
 It's just been our interview scheduling. We're not aligning with the stars with the security. However, Isom's, the process has allowed me to kind of get a much better idea of kind of understanding the third party space that comes with these integrations.
 So for those, if you are here, if for those of you who are just beginning to explore nervous scheduler or you've been hesitant, I can tell I see why.
 I mean, I love it. We've been trying to do it, but from a secure perspective, a lot of IT groups are a little nervous.
 But we did hear from Isim, they were putting and writing saying, they're looking for the end of this calendar year to sort that out.
 They're in, they're having an active project where they're working directly with the vendor chronify, they're the one who does the heavy lift for the integrations and the connectors for a new schedule to actually grid to make it more IT security friendly, not ask much stuff.
 So there's a light in the tunnels I shared with my team. I was like, we're keeping that door open. And so as soon as we hear that's been updated and revised at the end of the year, we're going to come back and take another look at it.
 And we're going to keep an eye on the enemy's good doing software. But it's good to know, it was a good process to understand that and to get it better.
 So if anybody is curious to see what they shared with me in terms of the details of the kind of security levels they're using, what they're going for, feel free to reach out.
 and I can share that and you can pass on to your IT teams or do your own evaluation. Yeah, thanks for sharing that.
 I heard a couple of wins there. One is the transparency and knowing what's going on, right? And two is knowing when to let go and focus your attention elsewhere, you know.
 Yeah, yeah, it's not always fun, but yeah, it feeds me up for other products. And now I'm gonna go back to my team and be like, hey, remember those that list of road maps that we want?
 Well, guess what, now we get to choose. What's our new adventure? Yep, yep. Thanks for sharing that, Greg. Cordelia, you had something to add?
 No, I have a new one that hasn't matured yet, but I expect it will, and I know I'm preaching to the choir here.
 But Vivian and Alex were kind enough after our last meeting I believe it was to walk me through a process of creating a mail campaign because we had lost our our town acquisition marketing person.
 And we need to mail out a bulk of emails to a lot of people. And so they were kind enough to just sit down and do a about 20 minute walk through and get me into a place where we are happy now.
 And we'll be able to meet our deadline before the end of this month to get our emailings out. Although, I expressed to them the fact that if we do multiple, you know, 50,000 emails, we are going to be Googleized and we're not going to be able, we might be put on a spam email alert.
 So I brought that to their attention. I think there's still one I'm gonna go through with it, but it's, anyway, and my point is that the value of this group, whether it's Alex and Kaelin and Vivian and all, or our members, it's worth the cost of admission.
 So, thank you all. Thank you. Thank you for sharing that, Crudeil. We did not pay Crudeil to say that. Uh-oh, and I also learned a new word, Googleized.
 I'm going use that and that is a thing maybe can you talk a little bit more about the email deliverability issue that we were discussing with Cordell?
 Just that when you're sending a very large amount of emails from your system if you're sending them in very large batches, Yahoo and Gmail and all of the places people do still have Yahoo email accounts are going to flag your platform's specific tag as a potential spammer.
 So you want to be judicious about how big your campaigns are. And I think it varies, but the guidance we were always given to advise clients on the ISIM side was under a thousand per individual email sent.
 So if you're sending a campaign out and it's got like 20,000 people, that's going to get a flag that it's spam.
 But if you're sending smaller snippets from a campaign perspective, you're less likely to get flagged as a spam bot. So you advise creating multiple email accounts to make that happen or how does that work?
 No, it's still going to see that it's coming from your platform because it's at isims.com and it's going to be yours.
 It's that I advise you to send smaller campaigns if possible. If you're going to run a campaign, don't run a campaign with 10,000 people in it.
 So if you're search results in the campaign that you're doing, and this is mainly in the Connect Engine, but also in CXM, and yields like 10,000 people, try to refine your search a bit better so that it's more targeted towards your specific role.
 I mean, in some of the broader roles where you're trying for a very broad pool, you're gonna want to possibly have multiple campaigns running at the same time in order to achieve that rather than having one great big campaign that's sending out 10,000 emails in a single clip.
 So I think you said one or 2,000 per person. 1,000 was typically the advice that we were given less than 1,000 people.
 Is that for a day or what's the time we do? Per, that would be out of your platform in a single day.
 And that's not all the emails that are going out of your platform. It's in a specific instance for a specific message with the same subject.
 Okay, good to know. Any other wins for the month of June? I don't know what this is, Terry, with the chief state bike drive you have.
 One thing that I am excited about, we just, I just finished sign off on testing for using an API to read, I'm going to call it read threshold jobs.
 And what it does is it goes in, it copies the existing job, because today we use evergreen. So it copies the existing job, moves candidates over and active bins who are no longer are no older than 90 days and close fills the old job.
 So, so it's going to eliminate our need for manual refreshing that we've been doing in the past.
 Well, if you've cracked that nut Terry, I mean we've been talking about that for six years. Yeah, let's go. Yeah, it's been tough.
 It's been tested. We can do it on a single job or we can do it on mass jobs. Wow, architecture of that a little bit more for everybody on the group.
 Are you going from isons to a middleware and a data warehouse where the conversion is happening and then sending or you go straight to your exam.
 So the vendor that we're using makes an API call to, well, we send it a job number. So let's say we're, you know, I've taken all 2,800 jobs that we have and I've assigned them one to nine randomly.
 And so let's say I'm going to refresh group 13. So I'll go to my console from the vendor, I'll type in 13.
 and what it does is it goes and it looks for all the jobs that are, have a repression number of 13.
 It then looks to see if anybody is currently an onboarding or the offer because it will skip those because we didn't want to mess with onboarding API calls.
 So it skips those and then ones that are good that don't have anybody in offer or onboarding. It then starts, copies the first job, moves people over in the same status that they were in in the old job.
 And it's it's basically like using a transparent activity. And then once it moves all the other people over and the copy of the people and the job is successful, it then lastly close fills the other job.
 If there's an error of any time, it closes the new job and leaves the old job up. So you said the vendor who's the vendor that you're working with?
 The cloud connectors work. They're now time to go. Okay. That's very cool. Yeah, it's it takes about seven minutes to run for 25 jobs.
 So it's not too bad. How are you accounting for time to fill if you're closing the original job? We don't we don't count for time.
 We don't. Time to fill is not a there's not a metric that we're concerned with most of the people you know we hired 24,000 hourly applicants a year and most of them are you know walk into the restaurant they apply and they're hired in the same day.
 Yeah so we do ours a little bit differently similar but different we retain the original job we unpost that and then we have a job folder called advertisement where we re, you know, we just copy the primary job post the advertisement job and the move candidates from the ad over.
 So if you're not capturing time to fill, it have to think through how you could do that if you still needed time to fill.
 Yeah, now the system is a couple of things that I didn't talk about the trace ability. So we have a, we created a new field called original RECID.
 And so the API is writing the copy reps I being on the new one. So if the RECID was 2020, you know, 2,532, and the new ones 2026, 20100, it's copying that old RECID onto the new red so we have traceability and then it's also doing exactly what the candidate transfer activity does and putting traceability
 on the candidate as well so we can see what job they came over and then it's recording all of their recruiting work flow history in a field area.
 Nina to your point and from what Terry just described I think this is possible if you would go with a similar setup, one thing that a lot of folks don't know is that the audit trail is consumable via API and so a lot of your time to fill your actual information about when actions have taken place can
 be sent to a data warehouse and that data warehouse can transform that information for you and then it can be sent back to Isom via either an API or Power BI dashboard or something along those lines.
 So your technical problem is solvable. There's a couple of different ways that you could potentially solve it. Yeah, the formula to, I think keeping the primary job open and then just doing an advertisement.
 Like that definitely works for us. We have the same thing with the extra field to say what, you know, this is an advertisement.
 What's the primary job? So we can do all of that. Our system definitely works for us. So I'm just trying to figure out how you would do via API simply, right?
 Because getting those developers time to program that can be, you know, can be a challenge, right? Because we're working on, they're working on some pretty heavy duty AI projects right now.
 And so it's hard to get their time to write API. Yeah, that's interesting. This is where a third party like harmonize or cloud connectors would be your better bet because they'd be dedicated to you specifically, but that's money.
 So, yeah, yeah. Terry, what justified the expense? Primarily in B and then they're changing, you know, removing gold posts that we have to try to hit.
 You know, they're not prioritizing jobs based on age any longer, they're prioritizing based on either sponsored or easy apply or apply network.
 And right now we still can't use the apply network because they don't indeed doesn't offer field swapping yet. So it doesn't look at and what that means is it won't look at our hiring organization from the portal.
 It looks like a hiring organization at the platform level. So you know for us and you know I don't know seems experienced it with their multiple reigns, but if we use the apply framework, every job we have, whether it's a fox concept or a North Italian, we'll say two-state factory as the hiring organization
 , that's supposed to be fixed by the end of June, so that we will be able to use the apply framework, but any other reason that just by this is we get many people that try to reapply to a job and because those jobs have been out there since 2020.
 It says you've already applied to this job and can't reify. So this is going to afford us the ability to have somebody feel to reapply or rehire us to come back in very easily.
 How many are you doing? How many jobs are you refreshing at a time just without setting off any alarms over there at our friends that indeed?
 On average there's about 24 jobs in each group and you can assume that somewhere between four and eight of those jobs will have somebody that is currently in onboarding, so we won't refresh those jobs, so it should be somewhere between 15 and 20 by the time.
 And they're random. They're not, you know, it's not like we're doing all the jobs at one location. They're random across, you know, I ran a job report And then told co-pilot to assign each of those jobs a random number between 1 and 90.
 And then re-uploaded the file so that they have a refresh number. let's see if I can. Wow. I'll show you, um, 50 seconds before it, I'll show you the screen real quick.
 So this is my refresh number 13 and you can see we added a field to omit them from the refresh.
 So if we don't want to request a job, we have the ability to keep it out, but then what it does is, oh, let me try and do this because those are some of those are my clothes filled.
 Is that a myth from refresh? Is that a manual field that may be the recruiter updates or how does that get populated?
 We bulked it to everything. So every job that wasn't recently refresh is going to be omitted from the refresh. So when I come in here, you'll see that you know, we have these 2026 jobs like this one was refreshed.
 So I can come in here with the details. And I can see that we we rewrote the externalizing. So this job was It repressed on 624 and it was 20, 20, 30, 4, 33.
 We keep the original wreck ID here, and then on the candidates, go to five. So it's on the person, not the recruiting workflow profile.
 Makes sense. So on the source tab, we capture their original application date. Now, you can see that they were transferred from a web service, and then, and it records all of their recruiting work for history.
 And if they had had multiple jobs, if this job had been regression multiple times, it would say, you know, 34, 33, and their work for history, and 19 to 20, and their work for history on that job.
 So it will continue to keep their recruiting work for history. It works very well and we're very happy to, like I said, I did final sign off on it this morning.
 And so they're going to build it in production. It should be in production by Tuesday. I have two questions to you, Terry, one, how long did this take and two, are you concerned about backlash from indeed?
 No, I'm not concerned about backlash from indeed. we're not repressioning a huge number of jobs every time and they've already flagged our jobs, but when we ask them what flagging our jobs means, they still can't tell us.
 So I'm not so concerned with indeed. We started that this is a considerably larger project than what RIT department was expecting.
 I tried to explain to him that, you know, think of this as onboarding, you know, but just a different, you know, it's that kind of project.
 It started in February, and we were supposed to go live in April, but because of, you know, some gaps, and we realized, oh wait, you can't see these fields, you need to see these, and we needed to add, you know, like some custom fields into both the job and the person and took him to a lab.
 So it's been a long project then. I will be thrilled to get it off my plate. Yeah, I bet. I mean, it's really, really a smart solution.
 Is this something that you came up and that you took to leadership and tried to persuade him it was worth doing or were you asked to do this?
 You don't mind me asking. No, I took it to leadership. I actually met with John who was harmonized HR originally And our IT department, you know, they paint who our vendors are for the most part.
 And they said we already have a relationship with Tavio for the cloud connectors. So, so this was something they had never done either, but they really did an amazing job.
 That's great. I mean, that's really the leading edge of this profession, where from your CUC opportunities for significant value creation and you know you have the relationships and you have the skills and knowledge and the insight you can move the needle so kudos well done.
 Thank you. It's been a long project that I am I'm happy to roll out live next week. And grads. Yeah thank you.
 I tell customers all the time that if you have enough time money and will almost every problem is solvable by API.
 And this is a great example of one. Yeah, it was because of an increased ad spend that we're going to have that we were going to have to do with Indeed.
 This is a huge cost benefit to us. Yeah, thank you. All right. Any other June wins and if not we'll open the floor to questions.
 All right, who has a question today? No question to you big of small. I'll ask him. There's something I, you know, and living in your article yesterday about the tokens and then seeing it in the summer release, you know, I had to ask our vendors because we also use the cloud connectors for a custom docky
 sign integration. And, you know, we have about 140 people a day complete that docky sign aspect of onboarding and so I was worried that each time they came in.
 It was it was a token, but they were able to confirm that it's only one token per day that they're using.
 So I was just curious if anybody else in this group is worried about the token limits on new integrations. They think can you quick recap of what we share?
 Yeah. So isems in the summer release release that coming this fall, they're going to start enforcing policy they've always had and that policy is that you get a hundred tokens per day for each existing integration.
 New integrations will get 60 per day. So a token is basically, anytime one system talks to another system, I'm going to butcher this if I'm trying to simplify it really hard, but anytime one system talks to another system, that system needs a foot in the So think about anytime you log into something
 using Face ID or something else and you have that little box at the bottom that says keep me logged in so you don't have to keep logging in.
 That's essentially a user level of creating a single token and a single session. So, ISM's integrations do the same thing, a poorly written integration will create a new token for every single session.
 And the amount of load that that puts on your actual system and their overall servers is much higher when it has to authenticate the session for every single action that takes place.
 So that's why it's a problem. And it's one of the reasons why they're enforcing it because global memory is getting very expensive, servers are even more money.
 And overall, this is a way that they can make sure that their systems are faster for all of their customers without greatly increasing their hardware costs.
 So usually, a properly written integration is going to be written in a way that it's only using one token per session.
 So it opens a token during the day. Does those 130 actions Terry just described? And then in the next day, it opens another token and for that single individual integration.
 That's a properly written one. So how do I know whether I have an issue, what are I going to do with you?
 You have to ask each one of your vendors if they're authenticating every session and if there's new token in their integrations.
 So this is one of the reasons is actually giving you several months before they enforce this rule because if your integrations, you need to know all of your vendors and just simply put out an email, they're gonna send you an email that explains the policy change, and they're basically asking you to share
 this with all of your vendors and make sure all of your vendors are compliant with it. That's very interesting. Is there a particular class of vendor that you think would be more likely to be an offender?
 Not necessarily. So Greg, who had to leave, had asked me in the thread about whether a flat file integration would use a token or not.
 And that's kind of a maybe yes or no, it really does depend on how the integration itself is written a flat file integration by itself should not use a token if it's a simple flat file.
 But if it's a flat file that's going to a third party and that third party is then sending information back in as a result of the information from the flat file, it could potentially be using one.
 Awath uses a totally different set of authentication and in some cases, depending on the system that's Awathing into ICIMS, it could be using tokens.
 So this could potentially touch just about everything. They are going to let, they made it very clear in some of the documentation that they said, they're going to let the worst offenders know if they've got integrations that are consistently going over this number.
 They're going to reach out to you individually and say, this integration is a problem child and you need to ask your vendor to rebuild it.
 But it's not a bad idea to get ahead of this. This is kind of one of those things where system administration at InfoSec kind of should be in sync.
 So I would share this with your InfoSec team, explain the different integrations your system touches, and ask them how they would approach this, because in theory your InfoSec team at most larger organizations should be aware of the kind of traffic that's happening between one system and another system
 and have some kind of documentation about all of your integrations. I don't see that often enough. But in theory, they should be informed.
 And if somebody answered going over to, the integration just stops, what happens? They are going to throw an error, a 409 error.
 So it will literally stop you from functioning. Wow, okay. Yeah. So coming this fall, So give you plenty of time to start talking to all your vendors and make sure they're not authenticated and you token with every action.
 They're just authentic and authentic and you token per session and those sessions are, you know, multiple actions taken within a single day.
 Great. Thank you. Uh, let's say hello to Leslie, Leslie Shore from Garment. You there, Leslie? I don't see her name.
 Yes, I am. Hello. Hi, Leslie. How are you? I'm good. Um, yes, I am new to Garment. I just started this week.
 I've worked with Isom's in the past. I was Isom's administrator at my former company, that just learned about this group and one of my co-workers suggested that I joined so they had a lot of good information so that's great are you are you are you feeling Amy Williams see maybe you don't know Amy she
 was she was at Garmin no I'm not familiar with her okay well welcome anything on your mind as if you get to learn Garmin's Isom existence?
 Considering that it's just my first week, not yet, I'm just learning how Garmin has their system configured and we'll go from there, but maybe future.
 Okay, great. Well, welcome and we hear every Friday. All right, it sounds good. Thank you. Absolutely. Absolutely. Hey, I did want to call out Amanda, did you get the answer that you needed for your question?
 I think a couple of people did respond, Cheryl included, but your question around deleting candidates and GDPR without having GDPR configured?
 Yes. Thank you so much for following up. I had a feeling it was as simple as you guys were confirming, so I appreciate everybody's input.
 Thank you. Okay. Let's see. Nine. What's going on in your world? Oh, we've got quite a few little enhancements to items, some cool templates that are new practice integration steam is using.
 Right now they're doing a lot of manual and so we're going to start tracking it. They're going to do some reporting.
 So they're pretty pumped about that and then and get ready to start that automation by September. We're gonna have some automation that when you update the status to submitted to hiring manager, it's gonna do something a little different for certain professions because our recruiting team is not gonna
 do end-to-end recruiting for like our front office and dental assistant jobs. So pretty excited about what the system is capable of doing.
 I presented that to our president this week or last week, and then just some other, you know, intake form, we do our, we have a custom tab on our jobs where we do an intake call with our hiring manager.
 And so we've added some new fields there and some customization there. And then our partner doctors, when we're recruiting for them, we need a little more information when the Harry manager submits the job.
 So we've done some customization to show or hide the fields with CFL when that specific type of job is selected.
 We couldn't go with the new job code because they still are in the same job code. So we had to come up with a creative solution to still get the information we need on intake.
 So lots and lots of items, things, lots and lots of UKG things and I've got about eight project requests right now for Isom's upgrades just you know small isom's upgrades so always fun all right thank you Cheryl what about you what's going on in your world What was that?
 Nice to know what's going on in your world. Oh, nothing much, honestly. I've been out of the office and traveling a whole lot this past six weeks, so I'm barely catching up.
 Where'd you go? Went to Vegas and died in the heat for a family reunion. reunion, but I've been in Texas for a few, like, about four weeks.
 So helping out my inlaw. All right. Well, also hot. Not suggesting anybody go there at this time of year. Yeah.
 Yeah. This is not the optimal time. All right. Any other questions for folks on the call? Just in general, I want to call everybody out.
 He left. So unfortunately, I can't get him to speak to it, but Greg posted something about a new New York ghost job bill that is on the docket and potentially going to be signed into law.
 And essentially, it's a doozy, just like the Ontario one we talked about a while back. So, keep your eyes on that and apparently there's one in DC and Massachusetts that are not far behind.
 So, if you are consistently posting positions for the sake of sourcing that don't actually have a position behind them, some of these laws could bite you as they're starting to be passed.
 And how would they prove that? You have to prove it. So that's the fun part about some of these laws.
 A lot of them are if you yourself, the company, you can't, a candidate files a complaint that they are, you've got to go as job that you posted.
 That complaint goes through whatever litigious body in the individual state puts together that you are the one with the burden of proof to say that this position is a real position and here is the headcount and all the different moving pieces associated with it.
 So how does this actually go down though? Because are there really candidates out there who are thinking wow this is a great opportunity to to sue a organization or is it law firms that specialize in this who find people it's a ladder it's ladder because the candidates don't get the money because otherwise
 oh my god yeah that would be awful yeah no but I got to say I remember when Colorado first put their pay transparency law into effect I believe they were the first a lot of candidates and people were just reporting companies left and right either because you know they got the client or this or that and
 so it was getting out of hand and I read that post from Greg and my first thought was it's five thousand dollars per posting per platform and I'm like what happens when you have jobs in our super super old because you guys all know like those random job those job boards that scrape other sites and then
 your job just sits out there but like all the information is not included in the job description. So it's like, that's, this is looking, yeah, and I, this is going to be bad.
 So honestly, I think that this is a good opportunity for you to start pushing back on indeed and zip recruiter and all the places that scrape because if you didn't physically post that job and no longer have control over it because it was scraped, now you're liable, potentially under some of these, even
 if it isn't something that is active out there. So I think that there's a layer of this where your vendors are potentially going to be the ones that aren't the problem because the job's not real anymore.
 Patrick said with this practically in the practice of evergreen repositions. Evergreens are problematic for a lot of reasons. In general, evergreens are not ideal, so I think that it's definitely going to put a lot of pressure on in documenting evergreens.
 And the reason that I say that is just trying to fill, I know was bringing it up, actually being able to kind of meaningful metrics out of your system with evergreen jobs is very difficult and it just makes that even more difficult.
 So the practice of stopping an evergreen within 30 to 90 days is going to get more important. So I don't necessarily know if it'll end it, but closing those evergreens consistently lean on a regularly scheduled basis.
 And then reposting is going to get even more important. So Terry's integration. Yeah. Terry, you're addressing laws. You didn't even know we're coming.
 Yeah. Very interesting. Patrick also said we may post a position that is pending customer approval. And then we make it a surprise that government funding didn't happen.
 Oh. Yeah. Wow. I mean, we run into that too. You know, if you're opening a new dental practice and something comes up with one of your inspections, I mean, you can't, you don't have a lot of control over when that practice is actually going to open and when you need to hire the person.
 And so hopefully intent will be considered, right? We can document what happened and why we didn't hire that person that that like is this a problem that we really want our legislator solving?
 Is this the best they can do with their time? Your lips. Yeah. I mean, it is a big problem from the perspective of fake jobs.
 I don't know if you've been on indeed searching for a job in any time recently. I will look at the things that my clients are posting just to get an idea of what's out there for them.
 Sometimes and there's a lot of noise out there that is questionable whether it's real or not. I think that's what this is trying to stop.
 But platforms like Indeed and The Rest are really the problem children when it comes to this kind of thing. So I think they're going out about it the wrong way.
 I think they should be looking at the job boards first and then the employers second. But I don't think that that's a nuance.
 I don't think the legislation is aware of. I'm on computer shopping this weekend. Get a Mac. I'm the Mac Nazi here.
 I probably should get a Mac. So anyway. All right, so I didn't get to tail into that conversation. Patrick also on board.
 You know, I have a PC for working, I have a MacBook Pro for my personal, and there was a two-week curve, learning curve, of me regretting having tried to make the platform switch because there were so many, so many jits.
 You lose sight of how much the tools of your trade become just something that is second nature to you. It also took a week for me to get all of my texts over I message for some reason, but I will say when I called Apple tech support, I had a question about the insurance policy because I'd accidentally
 purchased the wrong one and the guy, I got a US domestic person in like two minutes and I said, can you now transfer me over to tech support because I have a tech support question?
 And the guy said, oh, that's me too. You hear me? What is this? A human being? What a concept? A human being who was cross-domain, like can talk to me about, it was, it was, it was really a phenomenal consumer experience from that perspective.
 Anyway, we have a few minutes left. Any other questions or thoughts for today? Hey, Alex, one other thing, I don't know if you talked about this, if you'll talk about this and a meeting that I missed, but Liam and I have got to go to the on-site, Isom's admin training, and I will speak for Liam, but
 obviously, but if I would just make a recommendation. When it comes up again, if you get the chance to go, I would highly recommend it.
 There is a lot of training that most of you are going to know already, which is fine, but the office hours that I got to spend with tech support directly just one-on-one with Michael Mayer who's a tech support person there was awesome.
 And then, you know, the networking got to spend the three days or two days with Liam and then meeting other system admin experts.
 It was that well in the office hours were well worth it. And then, of course, the roadmap to which was, which was very cool.
 So again, I didn't know if y'all talked about the onsite or not. I'd seen some people were questioning whether they should go and I would just say that I would recommend in the future if you want to go you should.
 So thanks for sharing that. Did everybody get office hours? Where's the only larger brand? We have to sign anybody who signed up for it.
 Really? Yeah, and it was supposed to be 20-minute sessions, and I spent nearly an hour in a weekly of Michelin's band even longer.
 Wow. Yeah, mine was about almost two hours with Michael Rizzo, who's one of the product managers there. And we just went through, we went through basically every tool we use in the ATS one after another.
 And I was like, this is what we're having trouble with. This is what we need. This is the enhancement we'd love to see.
 And now we're getting like multiple calls with like the subproduct managers within those tools from that original conversation with Michael Brizo.
 So it was incredibly fruitful for us, just like it was for Terry. I think I didn't know it was going to be so introductory training focused.
 So that part wasn't really super fruitful for me, but I gave the feedback to our account team saying if we could have an on-site session in ISMS where it's smaller group workshop-type things where we're just with the product teams instead of like an orchestrated training that would be even more beneficial
 for us. So hopefully in the future we see something like that but it was it was very cool. So Mike Rizzo was my CSM seven or eight years ago, when I first implemented Isems and he was a TSE, then a CSM, and now he's a product manager.
 Yes. And that is like the ideal arc for somebody who really knows Isems. He's been there 11 years. So I'm glad you got that spend that time with him.
 Yeah, he was great. Yeah, yeah, really cool. All right, guys, let me just say one thing. Caitlin, a couple of weeks ago, asked for member feedback in Circle and I just want to call that out.
 if there's anything we can do that's you want us to do or any other feedback that you have for us just go ahead and respond there she's only gotten one response which is what i'm calling it up thank you all right good to see everybody hope you have a restful and restorative weekend and next week is i
 always mess this up because i am bad with dates next week is fourth July July weekend so we will be back the following week july 10th and i hope everybody enjoys you forward to July and we'll see you next time.
 Bye all. Bye bye. Bye everyone.