Go founder-mode on recruiting

Recruiting runs on scattered tools and stale data. Superlink is one system. Point it at a role, teach it your bar, and it works the market with you. You keep every judgment call.

Anthropicclaude-fable-5
Built with frontier teamsA*SierraTuringPerplexityInception
  • post the req and prayIf they exist, we find themDeep people search that doesn’t stop until the talent pools are exhausted.
  • one search at a timeScale without losing the barSeven parallel searches used to require a team. Now it’s just you.
  • thirty screens a weekNo more interviewing noiseYour builders get their interview hours back for people worth meeting.

Claude for recruiting

Context on the left, agent on the right. Sourcing, scoring, and outreach happen in one flow, so you never leave the window.

your pipeline
the agent
Where do we stand, and what’s next?
Reading the pipeline · 18 candidates
Checking replies & scheduled follow-ups
Since you were last here: of the 5 contacted, Maya Okafor (9.2) replied and wants to talk this week. Dana Kim opened twice but hasn’t answered, so her follow-up goes out tomorrow. 6 new candidates landed in triage, and two look strong against the v4 rubric: zero-to-one payments profiles. The 4 shortlisted are ready for first outreach. My plan: clear triage, then draft the 4. Start?
Yes, let’s do it.
Scoring 6 in triage against the Eval Framework
Drafting outreach for 4 shortlisted
Drafts are ready, LinkedIn plus email, in your voice. One approval and I’ll enroll the 4 shortlisted.
LinkedInoutreach campaign · from Alex
approveview
Head of PaymentsAsk a follow-up…Anthropicclaude-fable-5

Evidence beyond LinkedIn

Most tools stop at the resume. Superlink follows the trail builders actually leave: from GitHub and Google Scholar to X threads and YouTube talks. All of it fuses into one profile, so the agent reasons from evidence of real work.

LinkedInGitHubpapersHugging FaceYouTubeX
Maya Okafor
Eng Director · Stripe
9.2
LinkedInGitHubpapersHugging FaceYouTubeX7 sources
Runs money movement at Stripe · took the ledger 0→1
Maintains a vector-search library · 4.2k stars
2 first-author papers on retrieval · 1 patent
Ships eval datasets on Hugging Face
Conference talk: scaling ledgers · 12k views
Writes threads on eval infra
mayaokafor.dev · talks + projects
likely openbuilder
6,081 known by 2+ of you
your team network
Alex ChenFounderLinkedIn8,940
Sam TorresHead of EngLinkedIn6,204
Ravi PatelInvestorLinkedIn4,867
Jordan LeeAdvisorLinkedIn3,401
One searchable network23,412 people

Plus every network you already have

Send one link and your founders, investors, advisors, and teammates pool their networks into the same searchable graph, credited to whoever knows each person. When a candidate clears your bar, you already know who can make the intro.

“Deep-sea robotics leader who loves the ocean, thrives in ambiguity, and is off LinkedIn. Superlink found them in a market of maybe fifty people.”
Sahil Abbi

From your head to a shortlist

You already know what exceptional looks like. Superlink draws it out of you, sharpens it into something the system can verify, and works the market until the shortlist matches it. Hours, not weeks.

open roles
Head of Payments Infra
Founding Designer
GTM Lead
drop a link to your JD
scorecard
0-to-1 infra
Payments depth
Startup pace
Scaled an org
calibrated on 12 profiles you rated
curated datasets
LinkedInPooled team networks23,412
GitHubPayments OSS contributors8,143
Google ScholarSystems conference authors4,867
XInfra discourse on X3,209
4 of 12+ sources39,631 people
warm shortlist
Maya OkaforStripe
connected to Sam Torres
9.2
Dana KimSquare
maintains a payments OSS library
8.7
Lucia ReyesRippling
authored the ledger-scaling talk
7.9
Superlink
Your last three hires share a shape:0→1 infrabig-co pedigreeshipped a ledgerWant me to weight it into the scorecard?
Yes, weight it in.
Scorecard re-weighted · v3
Superlink
Done. One catch: this bar now screens out founder-shaped profiles, and a third of your current team came from early-stage shops before you hired them. Keep founder profiles in scope?
Good catch - those are some of our best people.
Bar locked · 9 candidates re-scored

It spars with you

Fully autonomous recruiting agents collapse on quality; Superlink stops for your judgment exactly where it matters. In calibration it scores real profiles, you rate its calls, and it pushes back: probing what the obvious hire looks like, and how far you’ll go for someone who doesn’t look the part. Every call you overrule sharpens the rubric.

The talent map

Before any search runs, we map the market with live research, your calibration verdicts, and the paths your own hires took. Each ring is a sourcing strategy, not a quality tier: its own keywords, exemplar trajectories, and scorecards. Hidden gems are real skill and experience transfers a title search would miss. The agent works from the map and iterates on its boundaries with you.

reachtargethidden gems
ring 1 · reach — the exceptional fewring 2 · target — right title, right stackring 3 · hidden gems — loud in code, quiet on LinkedInmap exhausted · 214 evaluated · 3 above your bar
map exhausted →86 evaluated · 2 above bar127 evaluated · 2 above bar214 evaluated · 3 above bartalent mapsearched ring by ringsearch passpass 1segment · bullseyekeywords41 foundpersonas28 foundlookalikes17 foundpass 2segment · hidden gemskeywords18 foundpersonas12 foundlookalikes11 foundpass 3segment · new datasetkeywords33 foundpersonas21 foundlookalikes9 foundevaluaterubric · v48.84.27.16.45.13.99.15.56.2yield check+2 above barsupply rich+0 above barsupply thinning+1 above barsupply spentbuild new datasetOSS contributors+8,143conf authors+4,867model authors+2,315feeds the next pass

Until the map is spent

Your calibration becomes a rubric the agent can check itself against, and then the loop runs: search a segment, evaluate everyone in it, keep what clears the bar. When supply thins, it builds new datasets and keeps going until the map is exhausted, on LinkedIn or off it. New frontier models slot straight in the week they ship.

Outreach that gets answered

A reply depends on the right channel, the right moment, and the right sender. Superlink runs LinkedIn and email as one sequence, drafted from the profile’s evidence so every note leads with what only a colleague would know. You approve before anything sends.

LinkedInLinkedIn inviteday 0

Hi first name, recent workcaught my eye. We’re building exactly that at company.

wait 2 days
Initial emailday 2
Quick intro, first name?

type @ to personalize
Maya Okafor
Head of Payments Infra · candidate
Meeting booked
Tuesday
Building payments at Acme9:02 AM

Sam mentioned the ledger rebuild you led. We’re at that exact stage, and he offered to introduce you

LinkedIn
Connected on LinkedIn
Thursday
LinkedIn
Would love to chat. Does Friday work?
Google Calendar
status → Meeting booked · Fri 2:00 PM
Friday
Granola
Call transcript granola
“open to the right zero-to-one”

It remembers every conversation

The email that went out, the LinkedIn connect, the reply, the calendar booking, the call notes: it all lands on one timeline per person, long before anyone touches your ATS. When someone answers, the sequence stops on its own. Five tabs become one.

Recruiting on the frontier

For the searches that decide your trajectory, we embed with you: a forward-deployed recruiter running the engine, custom datasets built for the role, comp and close through the signature. We recruit every layer of the frontier, from the substrate that powers compute to the products that put models in millions of hands.

The frontier talent stack, from substrate up through distribution
Distribution

Product surfaces, go-to-market, and delivery that turn model capability into products people use.

Systems

Infrastructure, safety layers, and orchestration that turn trained models into production services.

Intelligence

The development loop that turns research, data, and compute into trained models.

Compute

Chip architecture, networking, and the toolchain behind usable processing capacity.

Substrate

Materials, manufacturing, and facilities behind AI: silicon to data centers.

Built by doing the work

We tuned every scorecard, segment, and sequence in the platform on live searches across the frontier. A few of the seats we signed:

REQ-031Sierra
Agent Engineer
a title that didn’t exist a year ago
Signed
REQ-017Turing
Research Engineer
every lab is bidding on the same few hundred people
Signed
REQ-023Wetstone
Deep-Sea Mining CTO
maybe fifty qualified people on earth
Signed
“RL engineers are hard to land. Superlink made the search and conversion look easy.”
Anshul Bhagi
“Superlink ran circles around our talent team for a hyper-specific CTO search.”
Gautam Gupta

Operator judgment, built in

We ran the hardest searches by hand with one question: how can AI make us superhuman? What we learned is how our platform works.

Shalin MantriShalin MantriCEO, SuperlinkDirector PM at Google, early PM on Uber ATG. Stanford CS.
Tyler LambeTyler LambeCTO, SuperlinkCoding since age 11. Built and sold tech-education bootcamps.
Peterson ConwayPeterson ConwayPaypal Mafia TalentTalent for A*, 8VC, and Founders Fund. Placed Palantir’s CTO.
Alex DworskyAlex DworskyRecruiting LeaderRecruiting at Uber, Tinder, and Bird.
Kaajal BahetiKaajal BahetiPrincipal PMProduct at Handshake, ex-Workday.

Run it yourself, or run it with us

Run it yourself

Platform

The full agent, run by your team: your networks, deep profiles, your calibrated bar, and outreach in one system. Point it at a role and it’s searching the same afternoon.

Networkwarm paths
Deep profiles12+ sources
Learning loopsencoded taste
Outreachin your voice
  • Pooled team networks, credited warm paths
  • Full agent: background sourcing & scoring
  • Calibrated scorecards & shared skills
  • Drafted intro requests & outreach
We recruit for you

Embedded

The same engine, run with our team: operators with deep networks of their own drive your search and hand you vetted, off-market candidates.

Shalin MantriShalin Mantri
Tyler LambeTyler Lambe
Peterson ConwayPeterson Conway
Kumar ChellapillaKumar Chellapilla
Jennifer AndersonJennifer Anderson
Thuan PhamThuan Pham
Akash GargAkash Garg
Wayne ChangWayne Chang
Asheesh BirlaAsheesh Birla
Kaajal BahetiKaajal Baheti
Alex DworskyAlex Dworsky
Eric RiesEric Ries
Shaherose CharaniaShaherose Charania
Kittu KolluriKittu Kolluri
Jairam RanganathanJairam Ranganathan
Indrajit KhareIndrajit Khare
Haider SabriHaider Sabri
  • Full-cycle searches, kickoff to signature
  • A forward-deployed recruiter on the engine
  • Custom datasets curated from 12+ sources
  • Backchannel intro requests
Ready when you are.