Upswell completely eliminates the inbound funnel, replacing traditional job posts and resumes with structured, double-blind passes.
There is no inbound funnel to manage. Instead of collecting keyword-stuffed resumes, founders explore live, authenticated engineering history. Discovery replaces pipeline triage.
Conversations require mutual structural fit — neither side can cold-pitch the other. Upswell stays quiet until a role and a builder line up, then automates the introduction.
The code speaks first. Names, employers, and GitHub handles stay private, so both sides evaluate on real output. Identities reveal only after a mutual match.
Build and Role map to the exact same six axes. The role describes the spec; the build is shaped by real engineering history — same structure, opposite ends.
Generate a mock role pass →Build Pass — what founders see
Real engineering history. Proof, not claims.
Role Pass — what builders see
The mission and the spec. Never a JD.
BL–ML7734 · Build Pass
Staff · 7 yrs experience
8.4
SIGNAL SCORE
BUILDER STACK
ACTIVITY · 22 WKS
CLAIMS
Shipped ML to prod
4 attestations
Built training infra 0→1
5 attestations
Open source ML contributor
3 attestations
Led ML team of 5+
2 / 3 · needs 1 more
SKILLS
BACKGROUND
Led ML infra at two early AI startups, both pre-Series A. Owned the full stack: data pipelines, training, eval harnesses, low-latency serving. Cut p99 from 800ms to 90ms on a team of three.
ABOUT ME
Most models die in notebooks. I build the systems that get them to users — and keep them there.
IDENTITY
CO–APZNPN · Role Pass SEED
Senior · Founding hire
◇ Save
$200–275K
Immediately · Remote
ROLE STACK
REQUIREMENTS
REQUIRED SKILLS
OVERVIEW
YC W24, $3.2M raised, 8 on the team. We build the model-training infra layer for next-gen ML pipelines — ingestion through low-latency serving.
WHAT YOU'D OWN
Training infra, eval harnesses, and the serving pipeline from experiment to production. Third engineer. Full ownership.
COMPANY
Same six axes — the role maps the spec, the build is shaped by real history. The match is the overlap.
hover either pass to bring it forward
Your pass builds itself directly from your live engineering history, tracking your real output so you never have to write another resume. Connect GitHub once to instantly visualize your true depth, while our double-blind architecture completely protects your identity.
Commit velocity — your sustained shipping cadence, naturally weighted to recency.
PR quality — the deep, collaborative footprint of your code reviews and merged work.
Stack depth — your proven execution and technical breadth across our six engineering axes.
Verified vouches — high-context recognition from peers who shipped code alongside you and can vouch for the work behind closed repos.
Founders run their pipeline from the hub. Builders manage their pass and incoming matches from the console. Same data, two lenses — never a stale resume in sight.
ML Infrastructure Engineer · Senior · Remote · $200–275K
Active matches
Ranked by signal · double-blind
Role Pass
What builders see
ML Infrastructure Engineer · Staff · Open to Seed–Series A
Incoming role matches
Founders who match your pass
Build Pass
As founders see it
Define stack requirements, project depth, and team mission in under two minutes.
Search builders by proven environments — early-stage experience, specialization, code-depth levels.
Read short, high-context passes with real GitHub metrics, verified vouches, and brief bios.
Connect with clarity. Identities unmask only when both sides mutually opt in.
Connect GitHub and select your background badges. Your pass updates from live data.
Stack radar, commit activity, and code depth populate automatically. No portfolio to maintain.
Past colleagues attest to specific shipped features directly on your pass.
Name, employer, and GitHub handle stay hidden until you and the founder both agree.
Recruiters optimize for placement fees. Job boards optimize for volume. Neither was built for a seed-stage founder hiring their first engineers.
No applications. No resumes. No cold outreach.