Call —
Transcript streaming
0:00 Caller: Hi. Hi. Who is it? It's just here.
0:00 You: This call will be recorded.
0:04 You: Hello, Theo.
0:09 Caller: Oh, cool. Chad, yeah. Yeah, yeah. I think I did I just mess you like a second ago on LinkedIn?
0:09 You: Hello, Theo.
0:10 You: I'm Jed, just responding to your in-mail from earlier.
0:19 You: I move quick.
0:20 Caller: Yeah, thank you. Thanks for getting back to me. I always appreciate it.
0:20 You: or I try to anyway.
0:27 Caller: Cool. Is it a good time to speak about this now then or?
0:30 You: It is. Do you have a moment?
0:32 Caller: Yeah, perfect. Yeah, no, of course, of course, of course.
0:33 You: Thank you.
0:36 Caller: So I'm not sure of you if you read or
0:40 Caller: description I sent you, I assume you have.
0:40 You: Thank you.
0:43 Caller: I'm recruiting for HG Cadillist.
0:48 Caller: It's quite cool, overall, I think.
0:51 Caller: Obviously, yeah, it's an AI-Iapply engineer.
0:54 Caller: Yeah, the way they can describe the role to me
0:56 Caller: is you're effectively doing a startup.
1:00 Caller: once every nine months to 18 months that's kind of the gist of it really so I mean
1:00 You: So I'm in process with a number of roles, actually.
1:07 Caller: yeah we'll see when you right now are you currently looking for a new role okay
1:13 You: Two of them are already PE firms who are adopting similar strategies.
1:18 You: So I'm curious to know to what extent.
1:20 Caller: And okay, I mean, if I would do some
1:20 You: a catalyst would consider my candidacy with a, if an offer is made a window of end of June or first
1:30 You: week of July. That's where I'm anticipating the other processes will end those that come
1:36 You: with offers. I know it's...
1:40 Caller: it is quite tight to be perfectly honest with you I I don't know I can I definitely
1:40 You: Bye.
1:47 Caller: can guarantee whether they they would do an offer that quickly it let's say if you
1:52 Caller: could interview which would probably the first stage would let's say let's say
1:58 Caller: you'd get an interview you'd get an interview you
2:00 Caller: early next week. I think there's at least four stages to the interview process for this role.
2:00 You: well i ask because the most recent p e firm compressed their process to three days from
2:08 Caller: So I don't know if you still consider that.
2:13 Caller: I think I think somewhat somewhat.
2:20 Caller: Fair enough. I mean, I can always ask. I can always ask.
2:20 You: what I assume was much longer to hit the deadline.
2:23 You: So I know you have the flexibility.
2:27 Caller: Yeah, yeah.
2:27 You: And if it helps make your case, feel free.
2:30 You: I'll pass along some links for them to direct their preferred AI agents
2:35 You: to ask any questions they have about my methods or their results.
2:40 Caller: Yeah.
2:40 You: So perhaps that will help them make a decision one way or the other.
2:46 Caller: Sure, I mean, there's no harm I'm asking, right?
2:49 Caller: So what's your kind of deadline again?
2:51 Caller: Let me throw this dinner.
2:51 You: So I'm targeting offers landing in the last week of June or the first week of July.
2:56 You: I recognize that's about two weeks away.
2:58 Caller: Okay.
2:59 You: I recognize...
3:00 Caller: Yeah.
3:00 You: it's tight, so I'm willing to do what I can to accommodate whatever evaluation they need.
3:03 Caller: Cool, cool. So, I mean, with this role,
3:06 You: And I'm also, after this call, I'll be responding to your in-mail with the resources
3:09 You: you need to also help them ask whatever questions they would have asked of my profile.
3:20 Caller: The kind of people that are really looking for people with, obviously, I think the core thing is, it's quite customer-facing, and people who have built customer-facing products from zero to one. Would you say that that kind of fits your profile?
3:20 You: We're going to be.
3:40 Caller: Yeah.
3:40 You: 0 to 1. I've built prototypes. I've built some unusual things in the document processing spaces of automations in terms of manually manipulating computers. So I have a very large toolbox of solutions that can
4:00 Caller: Yeah, I mean, definitely that stands out on your profile is more, I think, to do you from other people that I've seen
4:00 You: be deployed across clients across industries. So I imagine that's why the other PE firms are
4:07 You: willing to move as quickly as they have and perhaps your client will as well.
4:20 Caller: get the roles at HG I think working in matter is definitely a big plus that I think
4:20 You: Thank you.
4:26 Caller: the new that they've got new CTO in recently and I'm getting the impression he quite
4:30 Caller: likes people who have worked for big names and stuff like that in terms in terms of
4:40 Caller: So, could you maybe go of your current role?
4:40 You: So the title is research and unfortunately
4:44 Caller: So on your recent LinkedIn, what's
4:50 Caller: you build a language of automation?
4:53 Caller: So what's, is we doing right now more research base?
4:57 Caller: Or because of it?
5:00 Caller: Yeah.
5:00 You: it's very misleading. What I do is I fill the gaps that represent manual work performed by
5:09 You: operational teams that benefit from engineering attention but are not wide enough to justify a full
5:17 You: engineering squad.
5:20 Caller: Yeah. Okay.
5:20 You: So a practical example.
5:24 You: Our go-to-market team is predominantly sales reps.
5:27 You: A lot of what they used to do is enriching profiles of prospects with relevant information.
5:32 You: I automated that away so they can spend much more time on the phone or actually engaging with the prospect versus learning about them.
5:40 Caller: Okay.
5:40 You: for our compliance and onboarding teams.
5:42 You: There's a very specific set of requirements
5:44 You: from our banking partner.
5:46 You: Compliance in the past had to open a dozen or so
5:49 You: different tabbed websites to look at the different pieces
5:53 You: of information about a prospect before they decide, accept,
5:56 You: or reject the application.
6:00 Caller: Mm-hmm.
6:00 You: Now they only really open four or five, and most of it is just summarizing data that I'm already collecting, or my agents are already collecting on behalf, or in the course of conducting research, find anomalies that require human judgment, taste, and accountability.
6:18 You: So a good example there is...
6:20 Caller: Yeah.
6:20 You: commerce client whose website has no checkout button. Those are usually the kinds of fraudulent applications we catch.
6:30 Caller: So I mean, in this, I mean, so they give me a couple questions to ask candidates.
6:36 Caller: One of the ones they've, they've, well, they want me to ask you is
6:40 Caller: And in what ways do you use LLMs to, oh wait, sorry, I'm going to pull it up, I wasn't prepared.
6:40 You: So a classic one is document
6:49 Caller: Yeah, it was, please describe and, oh, no, how have you implemented AI tools and LMs into your workflow?
6:58 Caller: Could you unsat for me if you could?
7:00 Caller: Yeah.
7:00 You: A classic one is document processing.
7:03 You: One function at Roe is their credit product,
7:06 You: but instead of using credit reporting agencies,
7:09 You: it creates a financial model of the customer
7:14 You: and then estimates the credit worthiness from that,
7:17 You: drawn from bank statements, transactions within...
7:20 Caller: We're going to be able to be.
7:20 You: the bank statements, submitted accounting statements, think income statement or balance sheet,
7:25 You: and then any other available information that can be found online or from our different vendors.
7:31 You: So all of that is ingested, made machine readable because prior to this, all the data was quite
7:39 You: literally.
7:40 Caller: Yeah, I mean, so, I mean, so, I mean, yeah, I think you're, I think you're, I think it's relevant.
7:40 You: hand-entered into Excel models, and then used to generate a representation of the client's
7:47 You: creditworthiness according to Roe's opinionated model.
8:00 Caller: I mean, my only concern maybe slightly is, I mean, so would you say you're quite hands-on then?
8:00 You: I am exclusively hands-on if I delegate it to agents, not people.
8:06 Caller: Because in terms of all the stuff, I'm excuse your hands-on, yeah.
8:12 Caller: Yeah.
8:15 Caller: Yeah, okay.
8:16 Caller: So, I mean, in terms of...
8:20 Caller: Have you quit your current role then, if you've been looking for new ones or what's...
8:20 You: I have not, but I anticipated the quarter over.
8:25 Caller: You have not yet.
8:26 Caller: You've not yet.
8:27 Caller: Is there any reason in particular looking, you're looking for a new role or?
8:27 You: So four weeks ago now, a startup founder reached out to me cold from my GitHub,
8:36 You: suggesting that I interview for a role that they have.
8:40 Caller: Yeah.
8:40 You: is still in flight but I anticipate an offer in the next week and a half or so
8:44 You: after the last round and so I think it was a good enough catalyst for me to see
8:46 Caller: Yeah.
8:48 Caller: Okay.
8:49 Caller: Okay.
8:49 Caller: And that's just kind of got your interest going in terms of...
8:56 You: well how else the rest of the market would value my accumulated skills now
9:00 Caller: Yeah, fair enough.
9:00 You: and experience.
9:04 Caller: So I mean in terms of Asia Catalysts, I mean, I think the whole thing really is they've obviously
9:10 Caller: building AI applications from zero to one for like enterprise scale, a lot of them like
9:19 Caller: Gen 6th.
9:20 Caller: I think your recent experience does look relevant.
9:20 You: So the role I'm interviewing for at the moment are paying
9:24 Caller: I mean, so in terms of the rate that I said to you and everything,
9:29 Caller: if you don't want me asking what package you're currently on,
9:34 Caller: is the one I've seen realistic?
9:37 Caller: Or?
9:40 Caller: Okay.
9:40 You: If it's all liquid, 450,000 plus total comp, if it is illiquid, so base plus RSUs or some illiquid compensation, the median there is 600,000 total comp.
9:53 You: So I imagine at least from what you'd shared, it's toward the bottom quartile, but I assume that's what...
9:55 Caller: Okay.
10:00 Caller: yeah I'm sure yeah um so yeah as i said it's it's yeah it's up to 250k um base um
10:00 You: they fought for the scope of the original role that they specced out.
10:03 You: And if we have that conversation, then I'm sure they can justify the outsized value I can deliver.
10:20 Caller: So, yeah, the average bonus is, as I said, as I put 100%,
10:20 You: Thank you.
10:22 You: Thanks.
10:23 You: Thank you.
10:24 Caller: it can go up to 150%, but it seems to me talking to them
10:25 You: I'm
10:30 Caller: that the 110 is kind of a standard.
10:33 Caller: And then obviously, if you're doing well,
10:35 Caller: then it goes up to 150%.
10:36 Caller: And then the equity in top, I'm not allowed to know the exact.
10:40 Caller: in the answer of it.
10:40 You: I don't think it's too far from the leading
10:42 Caller: But I'm led to leave it, it's quite, it's quite, it's quite extensive as well.
10:48 Caller: So yeah, it probably would be towards the bottom end, but I'm sure, I'm sure you could,
10:53 Caller: you know, if you're roughly happy with that, then we can go from there.
11:00 Caller: Okay.
11:00 You: process I have to exclude them. My only concern is if they're willing to compress
11:01 Caller: Okay.
11:02 Caller: Cool.
11:03 Caller: Yeah.
11:04 Caller: So, I mean, the only thing I need from you to really get it going is one, I need your CV.
11:07 You: their process and sometimes they're not and that's okay. You have the research, you have the
11:18 Caller: I don't know.
11:20 Caller: Oh yeah, oh yeah, I've just seen that, yeah.
11:20 You: resources, the links that I provided so that they can direct their agent at it.
11:26 You: Excellent.
11:27 You: So there are also a set of things, direct their agents to ask whatever questions of my methodology
11:32 You: in terms of agentic engineering.
11:34 You: And my GitHub is also there, so they're welcome to see what I'm up to and where I'm doing it
11:39 You: publicly.
11:40 Caller: Sure, thank you very much for saying that.
11:40 You: How long it was
11:43 Caller: That will help a lot.
11:44 Caller: The other thing I need from you, which I just sent over, is if you could write out to the
11:50 Caller: best of your ability the answers to these questions, I just asked you previously.
11:55 Caller: That would be great because they've just asked all candidates to answer them.
12:00 Caller: I just think, it doesn't have to be like an essay, but people have generally done around like a page or so, what I've seen before, but it's really up to you and how much detail you want to go into.
12:00 You: do you need? I guess how long of a response will they read?
12:18 Caller: I think then we'll just look, I think.
12:20 Caller: This is just kind of their way of screening out people who maybe won't be good for the role.
12:20 You: i'll have the
12:26 Caller: But if you want, I can ask first about if it wouldn't even be possible.
12:33 Caller: Or I can put your profile with the questions.
12:38 Caller: I don't know what's better.
12:40 Caller: Yeah. Cool.
12:40 You: I'll have the response to that question written up shortly and then just run in parallel as much as you can.
12:47 You: If it works out, great.
12:49 You: If it doesn't work out, then my next suggestion would be to go to the AI Tinkerer's website and look at their recent hackathon winners in New York.
13:00 Caller: Yeah, what are you, are you on there?
13:00 You: I'm not an organizer.
13:05 You: I'm not the owner, I'm not the organizer, I happen to know who the owner is, but that is where I direct recruiters whose roles are, for whatever reason, not a fit for me, but they still want LLC.
13:20 Caller: Okay, thank you.
13:20 You: I'm AI expertise because plenty of link plenty of pink people on LinkedIn claim to be experts
13:25 Caller: Okay, thank you.
13:26 You: and whenever I talk to them they tend to melt so at least here you have a ready source that
13:33 You: in my opinion hasn't yet been tapped of legitimate practitioners in the space
13:40 Caller: Sorry, can you pre-dust me again?
13:40 You: Of course.
13:42 You: I'll pass it along in the in-mail, so you'll have it not to worry.
13:44 Caller: Oh, cheers, I appreciate that a lot.
13:48 Caller: I mean, yeah, that's kind of easy information from you.
13:52 Caller: Just, yeah, I'll wait for you to respond to those questions for me,
13:54 Caller: and then I'll go from there.
13:56 Caller: And then if that time scale doesn't work, then I'll like
14:00 Caller: you know anyway what's going on. Is that OK? Cheers. Well, thank you so much for the call, Jed. I really
14:00 You: Splendid. All right. Excellent. Thank you, Theo. And then, of course.
14:07 Caller: appreciate that. And I'm looking forward to the response. Cheers, take it. Bye.
14:10 You: All right. Take care, Theo.