Employer beta opening Summer 2026 · Limited founding partner seats240+ companies on the waitlist · YC + a16z portfolios already in
For employers · Beta · Summer 2026

See past the résumé.
Vet with your AI agent.

Connect Claude or ChatGPT to Moss. Describe the role in plain English. Your agent queries 500+ verified graduates — real PRs, reviews, and interview transcripts — and hands back a shortlist you already trust.

500+verified graduates
3–5candidates per shortlist
0job postings, ever
Claude · Moss connectortrymoss.ai/c/alex-chen-ai
Find me a strong full-stack engineer who's actually built with LangGraph. Mid-level, US-friendly hours.
Querying Moss for AI engineers · 6+ wks shipped · LangGraph evidence…
Found 14 matches, narrowing on tradeoff reasoning & PR review depth.via mcp · moss.search
Top 3, with their actual work, please.
Alex Chen — top 5% eval, multi-tenant pgvector RAG (PR #2), 3 design interviews passed. Gap: light frontend.
Maya Roy — LangGraph orchestration on her residency capstone, defended live.
Devon Lim — strongest async writeups in cohort. Already passed Moss first-round.3 of 14 · async interview transcripts attached
Reply to Claude…
Why we built this

Hiring is a sourcing problem.
Every other tool solves the wrong one.

You don't need another ATS, another cold-outreach platform, or another AI sourcer that scrapes the same LinkedIn profiles everyone else is scraping. You need a talent pool with verified evidence — and an agent that can read it.

The old stack

Resumes, LinkedIn, ATS, recruiters

  • 01You write a job post. It sits on five boards. 400 keyword-matched applications come back.
  • 02A recruiter screens on degree, company logos, and years of experience.
  • 03By the time someone interviews, the only signal is a 30-min call & a polished CV.
  • 04You hire on vibes. Two months in you find out they can't ship.

Moss

Verified work, queried by your agent

  • 01You describe the role to Claude or ChatGPT. No posting, no ATS.
  • 02Your agent queries 500+ graduates — every PR, review & deliverable already on file.
  • 03Moss runs structured async interviews with the top matches before you ever see them.
  • 04You meet 3–5 humans you already trust. One conversation, one offer.
How it works

Hand the funnel to your agent.
Keep the final conversation.

Moss is built around a simple bridge: a connector your AI tool plugs into, the same way it plugs into Gmail or Notion. From there, sourcing, screening, and async first-rounds happen without you in the loop.

Step 01

Plug in your AI

Connect Claude, ChatGPT, or Cursor to Moss in one click. Same as adding any other connector.

→ moss.connect() scopes: search, interview ✓ connected
Step 02

Describe the role

Plain English. No req-form. “Mid-level full-stack with LangGraph in production, ships fast.”

you › mid-level eng, LangGraph, US hours, open to switchers
Step 03

Moss screens async

Our interviewer runs structured first-rounds with the top matches. Transcripts come back to you.

AAlex Chen9.2
MMaya Roy8.9
DDevon Lim8.4
Step 04

You meet 3–5

Your shortlist arrives in your inbox with full transcripts, work samples, and gap notes. You take it from there.

3 candidates pre-vetted transcripts ready first call: this week
The bridge · Built on MCP

One connector.
Any AI tool you already use.

MCP is the standard that lets your AI tools talk to outside services. Connect Moss once, and Claude, ChatGPT, or Cursor can search the talent pool, read interview transcripts, and pull verified work — all from the same chat you're already in.

Search the pool from your chat

No new app to learn. Ask your AI in plain English. Get a ranked shortlist back — names, fit, and evidence — without leaving the chat you already use.

Claude · with Moss
Find me a mid-level engineer who's actually shipped AI features. Open to switchers.
14 matches · ranked by fit
AAlex ChenAI Eng · Residency IV9.2
MMaya RoyFull-stack · Residency III8.9
DDevon LimAI Eng · Residency IV8.4
Available inClaudeChatGPTPerplexity

Interview the credential

Every resident has a live profile your AI can interrogate. Ask anything — about decisions, gaps, working style — and get an answer backed by their actual work.

Alex Chen · live profile● available May 1
Top 5% eval3 design interviews passed6 reviewed PRsGap: light frontend
How does Alex think under pressure?
Strong fit. In her week-5 design review she changed her chunking strategy mid-discussion after a mentor pushed back — defended the new tradeoff cleanly. Tends to over-explain when nervous; gets sharper with seniors in the room.grounded in 6 reviews · 3 transcripts
Backed byreal work, real reviews, real interviews
What you actually get

Behavioral signals.
Not keywords.

Every Moss resident builds in public — through a 4–10 week residency where every check-in, PR, design review and interview is captured. Your agent can query any of it.

Real work record

PRs, code reviews, architecture writeups, design reviews. Every artifact a working engineer would leave behind — and yours to query.

example:Show me her last 3 PRs and one self-review where she changed her mind.

Consistency & growth

Daily standups, week-over-week velocity, gate scores. The shape of how someone actually works — not a polished snapshot.

example:Did his output improve after the week 4 mentor feedback?

Async interview transcripts

Full first-round transcripts run by the Moss interviewer agent. Read the actual words, not a recruiter's summary.

example:How did she explain the chunking tradeoff to a non-technical stakeholder?

Verified gates & scores

Every resident clears scored gates set by working operators. Top-decile, top-quartile, on-trajectory — and the rubric is open.

example:Filter for top-decile system-design gate, intermediate level.

Mentor & peer signal

Notes from working operators who've watched the resident ship for weeks. Not LinkedIn endorsements — actual review notes.

example:What did her week-6 mentor flag as her biggest strength?

Live availability

Status updates from day one to today. See who's actively interviewing, who's heads-down, who just opened to offers.

example:Only show candidates open to onsite in NYC, available May 1.
The credential

See the work,
before the first call.

Every shortlisted candidate comes with a credential like this one. One link, one read — you'll know how they performed, what they shipped, and where they need support. No more gambling on a polished resume.

What you seeTheir performance band, completion record, and time-to-finish at a glance.
What's behind itSix weeks of reviewed work, scored by operators in the field.
How to use itShare the link with your hiring manager, or hand it to your AI to dig deeper.
Move your cursor over the card
m
moss.
Verified
This certifies that

Alex Chen

completed the AI Engineer Residency
Top 5%
Eval score
6 / 6
PRs merged
5 wks
Completion
Verify or interview this candidate
trymoss.ai/c/alex-chen-aiISSUED 04.2026
Pricing · Founding partner pricing locked for life

One agent seat.
Unlimited shortlists.

You pay for the agent, not per-applicant or per-hire. The first 50 founding partners lock these rates forever — and shape what we build.

Solo founder

Pilot

Freefirst 30 days · beta

For founders making their first 1–3 hires. Hand the funnel to your agent and meet 3–5 vetted candidates per role.

  • 1 AI agent seat (Claude or ChatGPT)
  • Up to 3 active role searches
  • Async interview transcripts
  • Email support
Join waitlist
Most signed upGrowth team

Founding partner

$1,200/ mo · billed annually

For seed–Series B teams hiring 4–20 people a year. Multiple seats, role templates, and a hand in shaping the product.

  • Up to 5 AI agent seats, any tool
  • Unlimited role searches & shortlists
  • Custom interview rubrics
  • Direct line to the Moss team
  • Founding partner pricing locked for life
Apply for founding partner →
50+ headcount

Scale

Customtalk to us

For talent teams running structured pipelines. SSO, custom rubrics, dedicated success, and bring-your-own-agent support.

  • Unlimited seats & agent tools
  • SSO + audit log
  • Custom rubrics & gate weighting
  • Dedicated partner team
Common questions

Fair questions before you give your AI a seat at the table.

01
What is MCP, in plain English?
MCP — short for Model Context Protocol — is the open standard that lets AI tools like Claude or ChatGPT plug into outside services. Think of it like USB for AI: connect Moss once, and any AI tool you use can search our talent pool, read interview transcripts, and pull verified work without leaving the chat.
02
Do I need to use Claude? What if my team uses ChatGPT?
Pick whichever you already use. Moss is available as an MCP server, which means it works in Claude Desktop, ChatGPT (custom GPTs & agents), Perplexity, and any other tool that supports the standard. We add new clients as they ship MCP support.
03
How is this different from an AI sourcing tool?
Most AI sourcers scrape the same LinkedIn profiles everyone else is scraping. Moss starts with a different pool: 500+ graduates who've spent 4–10 weeks doing the actual work in front of operators, with every artifact captured. The AI isn't guessing from a CV — it's reading verified evidence.
04
Are residents really hireable, or are they all juniors?
A mix. About 40% are career switchers with prior senior experience pivoting into a new field, 35% are mid-level engineers leveling up, and 25% are strong new grads. Filter by level, gate score, and trajectory — your agent sees the whole distribution.
05
Who runs the async first-round interview?
A Moss interviewer agent, trained on rubrics co-designed with operators in each track. It asks structured questions, follows up on weak answers, and produces a transcript plus a scored summary you and your agent can both read. You can also bring your own rubric for founding-partner tiers.
06
What does it cost to actually hire someone?
Nothing extra. Your agent seat covers unlimited searches, shortlists, and interviews. We don't charge per-hire, per-applicant, or take a placement fee. Residents are paid through their residency stipend & income share, not by you.
07
When does the employer beta open?
Summer 2026. We're admitting founding partners in waves of 25, starting with AI-native startups and YC/a16z-portfolio teams already hiring AI engineers. Joining the waitlist gets you early access & locked founding-partner pricing.
Join the waitlist

Be one of the first 50 teams to hire this way.

Founding partners get early access, founding-partner pricing locked for life, and a direct line to shape what we build for hiring teams.

  • Beta access by July 2026
  • Founding-partner pricing locked for life
  • Concierge onboarding for your AI agent setup
  • First seat at our quarterly partner roundtable

Apply for the employer beta

We review applications weekly. AI-native teams hiring engineers are admitted first.

We'll get back to you within 5 working days. No marketing emails, no public list.