Follow-up · Oct 29 callCompetitor talent flow, in Claude

Start the Boltz conversation with their rivals' talent flow.

Boltz is a small team. Its founder can't see who Chai Discovery, Latent Labs and Genesis Molecular AI hired, or who they lost. We mapped all three over 18 months: 86 people joined and 48 left, every row read from a live profile, for 850 credits in total. Below is the report, what it cost line by line, and the recipe to run the next one from your own Claude.

See the Boltz report ↓ How it's built
Every number is from a live Crustdata pull · Oct 29, 2026
86
people joined Boltz's three closest rivals since May 2025
48
people left them, and we know where 39 went
850
credits for the whole three-company report
22
talent and people-team hires at 13 AI-bio companies since May
What I heard

Bring a founder something they can't get on their own.

If I can go to the founder of an AI bio company and immediately give them a talent flow report of one of their key competitors, that for me would be really powerful.
My business development strategy is all about creating conversations, not pitching people too fast.
Talent market reports on the top five AI bio employers across the UK, Europe and the US, and a talent flow chart out of that.
What it has to do

Four things a Talvo talent report has to get right.

The reader is a founder or a head of talent at a 30 to 200 person AI-bio company. They give it a minute.

01

Read cold in a minute

Joined, left, net. Where people came from and where they went, by type of employer. The few moves a founder would actually talk about.

02

Small snapshot, small bill

A rival Boltz's size should cost a known, small number of credits, so a report per target account is easy to budget.

03

Built by Talvo, in Claude

The same connector you moved to on the call, one ask, no growth team in the loop. The recipe is further down.

04

Any company, any region

UK, Europe and US, seed stage to Isomorphic size. Rivals for one founder, or the top employers in a region for a market report.

The Boltz report · May 2025 to Oct 2026

Boltz's three closest rivals, and who they hired.

Boltz makes open models for how molecules bind (Boltz-1, Boltz-2, BoltzGen) and sells a design platform on top. We picked the three companies whose product and buyer overlap most. Companies and titles only. Names come back in the pull.

Chai Discovery

The same product, a year ahead

Open structure model (Chai-1), then de novo antibody design (Chai-2). Pharma platform deals with Eli Lilly and GSK. $400M Series C in July 2026.

Latent Labs

The same design pitch, in London

Generative models for antibodies and peptides (Latent-X2), sold as a platform. Most of its team sits in London, like most of Boltz's.

Genesis Molecular AI

The same small-molecule bet

Pearl, a protein-ligand co-folding model, against Boltz-2's binding work. An expanded Incyte deal in May 2026. The biggest team of the three.

Left out on purpose: Isomorphic Labs (already in the Oct 27 report, and 440 profiles is a different league) and EvolutionaryScale (its team moved to Biohub).

Chai Discovery · San Francisco

Hiring like an AI lab, and almost nobody leaves.

  • 28 joined and 3 left in 18 months. Two of the three left to start their own companies.
  • Only 8 of the 28 came from biotech. 14 came from tech, AI and finance: two from Stripe, plus D. E. Shaw, Virtu, Nuro and Imbue.
  • 16 of the 28 joined into ML and engineering seats, most as "Member of Technical Staff".
28joined
3left
+25net
18→42profiles, May 2025 to Sep 2026
Notable joiners6 of 28
Title at Chai DiscoveryFromWhen
Security Lead, Platform & ProductsMinimalSep 2026
Member of Technical StaffStripeAug 2026
Member of Scientific StaffSeismic TherapeuticJul 2026
Member of Technical StaffGenerate:BiomedicinesMay 2026
Member of Technical StaffRaidiumApr 2026
Forward-deployed ScienceSeismic TherapeuticApr 2026
Who left, and where to3 of 3
Title at Chai DiscoveryToWhen
Founding ScientistStealth StartupDec 2025
Scientific ConsultantYogutideNov 2025
DevOps SpecialistJack & JillApr 2026
ML & engineering 16 in · 1 outScience & research 4 in · 2 outTalent & people 3 in · 0 outBusiness & ops 5 in · 0 out
This flow: 139 credits · every row read from a live profile · read Oct 29, 2026

How we read it: live search for everyone who joined or left since May 1, 2025, a check of our people index for anyone live search missed, and a live read of each of those profiles. Interns, internal moves and board seats are left out. "Profiles" counts people in our index, not company headcount.

Competitive intelligence

Where Boltz stands against the three.

The page a founder reads after the flow. Public facts only, each from the company's own announcement or trade press.

BoltzChai DiscoveryLatent LabsGenesis Molecular AI
Founded2024, a public benefit company since Jan 2026202420232019, as Genesis Therapeutics
What it shipsOpen models Boltz-1, Boltz-2, BoltzGen. Boltz Lab, with design agentsChai-1 structure model. Chai-2 for de novo antibody designLatent-X2, antibody and peptide designPearl, protein-ligand co-folding. Its own drug pipeline
Latest money$28M seed, Jan 2026 (Amplify, a16z, Zetta)$400M Series C, Jul 2026$50M in total, announced Feb 2025Expanded Incyte deal, May 2026, with an equity investment
Pharma and lab partnersPfizer (Jan 2026)Eli Lilly (Jan 2026), GSK (Oct 2026)GenScript, wet-lab partner (Jan 2026)Incyte
Where the team sitsUK, 17 of 28 profilesSF Bay Area, 39 of 42UK, 22 of 26California, 109 of 158 (59 in San Diego)
Profiles, May 2025 → Sep 202619 (Jun 2026) → 3318 → 4219 → 26134 → 158
Talent flow, 18 monthsNot run, on purpose+28 / −3+18 / −6+40 / −39
Sources: BusinessWire, The New York Times, Fierce Biotech, GEN, BiopharmaTrend, The Pharma Letter and company press releases, read Oct 29, 2026. Team location and profile counts are from our index the same day.
The read

Chai wins the AI talent fight

Chai pulls engineers from Stripe, Nuro and quant funds, and keeps them. A founder hiring the same ML profile is up against a $400M round.

The read

Genesis is the one leaking

39 left in 18 months, nine of them to AI labs and big tech. Its senior ML and comp chem people are the most movable bench of the three.

The read

Latent is the London rival

Same city, same DeepMind and Oxbridge pool, a team 26 strong. Every senior London hire Boltz makes is one Latent was also chasing.

The London angle

Boltz and Latent Labs fish in the same London pool.

Boltz reads as an MIT company. Most of its people are in the UK.

17 of Boltz's 28 profiles in our index sit in the UK. Latent Labs has 22 of 26 there, and it hires from the same places: Google DeepMind, Oxford and Cambridge spinouts, London AI startups. A Talvo report that shows a Boltz founder who Latent hired this year, and from where, is a map of the people Boltz will compete for next.

  • For BoltzLatent's 18 joiners by source, and the 6 who left
  • For TalvoA London story, told by a London firm
  • Next pullEvery ML and comp bio person in London who left a DeepMind or AI-bio seat this year
Share of profiles in the UK · Oct 29, 2026
Latent Labs22 / 26
Boltz17 / 28
Chai Discovery2 / 42
Genesis Molecular AI0 / 158
Four ways into a founder conversation

The Boltz report, plus three more Talvo can send.

Ordered by how fast you could run them. Play 01 is the report above, as a repeatable ask. Companies and titles here; names come back in the pull.

01

Hand any founder their rivals' talent flow

The askRun it this week

The Boltz report is the template. Name a target account, let Claude pick its three closest rivals, and get joined, left, where from, where to and the moves worth mentioning. A rival with 26 to 42 profiles took 87 to 139 credits. A 158-profile one took 624, because more people moved.

How it runs
  1. Pick the rivals and resolve them, free: company_identify
  2. Live search for joiners and leavers, then check the index for anyone it missed: person_search_live person_search
  3. Read the missed profiles live and group the moves: person_enrich_live

A founder can see their own hires in their ATS. They can't see the other side's.

→ A page or PDF for the founderOn demandA target account
What each rival cost to mapOct 29, 2026
RivalProfilesMovesCredits
Latent Labs262487
Chai Discovery4231139
Genesis Molecular AI15879624
Cost follows the number of people who moved, not the size of the company. The breakdown is in "What it cost" below.
02

The round closes, the talent team follows

Only Talvo can run this oneSignal · new talent hire

Since May 1, 22 talent and people-team hires started at 13 AI-bio companies. Several came weeks after a round: Superluminal Medicines raised a Series B in September and hired a VP and head of HR that month. Basecamp Research raised a Series C in September and added two talent seats, one through an outside talent firm. BigHat Biosciences raised a Series C in September and hired a talent advisor in October. Isomorphic Labs, which raised in May, added five since then.

How it runs
  1. Every Monday, list AI-bio companies with a seed to Series C round in the last 90 days: company_search on funding.last_fundraise_date
  2. Find talent, recruiting and people titles that started there since the round: person_search
  3. Leave it running, so a new match arrives on its own: watch_discovery_create

Talvo's Embedded partner fits the weeks between the round and the first in-house recruiter. This shows you that window, and when it closes.

→ Claude, Monday morningWeeklyNew round or new talent hire
Talent and people hires since May · 10 newest of 22Oct 29, 2026
Company · recent roundTitleStarted
BigHat BiosciencesSeries C · Sep 2026Talent AdvisorOct 2026
Isomorphic LabsSeries B · May 2026Talent Attraction & SourcingSep 2026
CradlePeople Operations ManagerSep 2026
IambicSenior Talent Acquisition PartnerSep 2026
Isomorphic LabsSeries B · May 2026TalentSep 2026
Superluminal MedicinesSeries B · Sep 2026Vice President, Head of Human ResourcesSep 2026
Chai DiscoverySeries C · Jul 2026TalentSep 2026
IambicSenior RecruiterSep 2026
Basecamp ResearchSeries C · Sep 2026Technical Talent LeaderSep 2026
Basecamp ResearchSeries C · Sep 2026Senior Talent Partner (Via Talentful)Sep 2026
Every row read from a live profile. Rounds checked against each company's announcement. Some of these may already know Talvo.
03

The top AI-bio employers by region, as a market report

Signal · team growth

The UK, Europe and US report, started. We pulled 284 AI-bio companies, ranked them by profiles added in the last 12 months, and hand-checked that each one builds AI models for drug or protein design. Isomorphic Labs added 164, Chemify 61, Cradle 35 and Superluminal Medicines 30. The next step is a talent flow on the leader in each region.

How it runs
  1. Describe the market in plain words and pull the companies: company_search in semantic mode
  2. Rank by headcount.growth_absolute.12m per region, and check each one
  3. Run the play 01 flow on the top three in each region: person_search_live

One report, sent to every AI-bio founder in a region, with their own company on the list or missing from it.

→ A quarterly reportQuarterlyRegion and segment
United Kingdom
  1. Isomorphic Labs440 now+164
  2. Chemify203 now+61
  3. Basecamp Research76 now+14
Europe, outside the UK
  1. Cradle134 now+35
  2. AQEMIA84 now+16
  3. Bioptimus43 now+8
United States
  1. Superluminal Medicines63 now+30
  2. Chai Discovery42 now+19
  3. Tamarind Bio21 now+12
Growth is profiles added in our index over 12 months, read Oct 29, 2026. Portal Biotech, PolyModels Hub and Living Models grew fast too, but didn't pass the AI-native check.
04

The talent seat just emptied

Signal · talent seat empties

Since May 1, 7 talent and people-team staff left 4 AI-bio companies. AQEMIA lost a talent partner and an HR business partner, and we found no replacement yet. Isomorphic Labs lost a talent director to Cohere, and a talent specialist went to an outside talent firm. BigHat Biosciences lost its head of people in June.

How it runs
  1. Watch the talent and people seats at your target accounts: watch_create
  2. When one ends, read the profile live to confirm the move: person_enrich_live
  3. Check whether a replacement has started yet: person_search

An open talent seat at a funded AI-bio company is the clearest Embedded pitch there is. Be first in, before the backfill.

→ Claude or emailDailyA talent seat empties
Talent and people exits since May · all 7Oct 29, 2026
CompanyTitleWent toLeft
Genesis Molecular AITalent and Competitor ResearchAlignityXSep 2026
AQEMIATalent Partner • Biotech & AINot listed yetJul 2026
Isomorphic LabsTalent DirectorCohereJul 2026
BigHat BiosciencesGeneral Counsel and Head of PeopleHyperfineJun 2026
Isomorphic LabsTalent SpecialistTalentfulJun 2026
Genesis Molecular AIRecruitingThinking Machines LabMay 2026
AQEMIAHuman Resources Business Partner & ProjectsNot listed yetMay 2026
Every row read from a live profile on Oct 29. Interns left out.
How it's built

The recipe behind the report, so the next one is yours.

No growth team, no manual work. The same connector you set up on the call, four steps, each search run once.

Step 1 · free

Resolve the rivals

Company page or domain in, company ids out.

company_identify
Step 2 · most of it

Live search, both directions

Current staff who joined in the window, and past staff who left.

person_search_live
Step 3 · cheap

Check the index for gaps

Anyone whose start or end date falls in the window that live search missed.

person_search
Step 4 · per profile

Read the gaps live, then group

Confirm each move, drop interns and internal moves, group by employer type.

person_enrich_live
01Rivals' talent flow
Using Crustdata: find the 3 closest competitors to [company page URL], meaning the same kind of product and buyer at a similar stage.
For each, build a talent flow for the last 18 months:
1. Live search for current staff who joined in the window, and past staff who left in it.
2. Check the people index for anyone with a start or end date in the window that live search missed, and read only those profiles live.
3. Skip interns, internal moves and board or advisor seats.
Group where joiners came from and where leavers went: AI labs and big tech, AI-bio and biotech, big pharma, universities, startups, finance and consulting, straight from school.
Give me joined, left and net, the 6 most senior moves each way (title and company only, no names), and the credits used per company.
02Round, then talent team
Using Crustdata: find AI drug discovery and protein design companies with 10 to 500 people that raised a seed to Series C round in the last 90 days.
At each one, list anyone who started a talent, recruiting or people role since the round, and anyone who left one in the last 6 months. Read each of those profiles live to confirm.
Show company, round and month, title, and start or end month. Flag the companies with a round and no talent hire yet.
03Regional market report
Using Crustdata, semantic company search: AI-native drug discovery and protein design companies in [UK / Europe / US].
Rank by profiles added in the last 12 months and check each one builds AI models, not services or hardware.
Give me the top 5 with their growth and latest round, then run the rivals' talent flow (prompt 01) on the top 3.
04Keep the cost known
Before you run anything: estimate the credits for this report from the live search totals (2 per profile) and the number of index-only profiles to read live (7 each). Show me the estimate and wait for my OK.
What it cost

850 credits for the Boltz report, line by line.

The figure promised on the call. Live search is 2 credits a profile, the index check is a fraction of a credit a row, and each profile read live is 7.

RivalLive searchIndex checkLive readsTotal
Latent Labs22 × 2 = 4425 rows · 16 × 7 = 4287
Chai Discovery20 × 2 = 4031 rows · 114 × 7 = 98139
Genesis Molecular AI103 × 2 = 206168 rows · 559 × 7 = 413624
The Boltz report2907553850
The lever

The live reads are most of the bill

Genesis cost more because 59 of its moves only showed up in the index, and each was read live to be sure. For a first look at a rival, skip step 4 and read those rows from the index for a fraction of the credits. Live-check anything that goes in front of a founder: an index row can say someone left when they haven't yet.

The rate, as discussed

Self-serve is open now. Doug is working on month-to-month.

Self-serve credits run 17 to 30 cents each. Direct with Doug, about 15 cents. Doug is sizing a month-to-month option around a handful of reports a month, instead of an annual plan.

Straight answers

The questions from the call.

Q1Is talent flow a separate product now?

No, it still runs on credits like everything else. What changed is the workflow: it's packaged so one ask in Claude runs all four steps, and the cost per rival is predictable.

Q2How were the reports Doug sent actually built?

With the same connector you now have, and a saved workflow that runs the recipe above. Nobody builds them by hand.

Q3How many credits does one report use?

87 to 139 credits for a rival with 26 to 42 profiles, 624 for one with 158. Open-ended exploring costs more because the same searches get repeated. The recipe runs each search once, and prompt 04 shows you the estimate before it spends anything.

Q4Do I need an annual plan to do this?

Self-serve is open today at 17 to 30 cents a credit. Doug is putting together a month-to-month option sized to this use.

Q5Can it tell me when something changes, instead of re-running it?

Yes. A watch on a list of companies or people, or on a filter like "new talent hires at AI-bio companies", sends new matches as they happen. You pay per new record it delivers.

What happens next

Let's get you self-serving this.

Everything below came out of the call. No dates we didn't agree to.

The short version

Their rivals' talent flow, ready before the first call.

87 to 139 credits for a rival Boltz's size, built in your own Claude, sent in Talvo's name.

Grab time with Doug Reply to Doug