Job Founding Full-Stack AI CRM @ Universal Hire nou

Senior Full-Stack / Founding Tech Lead (AI CRM)

What you’ll need (skills & experience):

• 3++ years building and shipping SaaS with significant 0→1 ownership.
• Strong React + TypeScript and modern API design; comfortable with relational data modeling, transactions, and indexing.
• Practical LLM integration: embeddings/RAG, tool-calling/structured outputs, streaming UIs; you can weigh model quality vs. latency vs. cost.
• Background processing: designing idempotent jobs, retries, schedules, and event-driven workflows.
• Security & privacy mindset: role-based access, least privilege, auditability, and basic threat modeling.
• Product sense & communication: you can clarify scope, trade-offs, and deliver shippable increments quickly.
• Tooling you’ll actually use (illustrative, or equivalent)
• Web app & APIs: Next. js/React + TypeScript for UI and server routes; tRPC or REST/GraphQL for typed contracts.
• Data: Postgres (Neon/Supabase/Cloud SQL/Aurora) + Prisma ORM; pgvector for semantic search; Redis (Upstash/managed Valkey) for cache/queues; object storage (S3/R2).
• AI: OpenAI / Anthropic / Google (plus OSS where it fits) via Vercel AI SDK/direct SDKs; LangGraph/LangChain/OpenAI Assistants for agentic flows where appropriate.
• Prompt Ops & evals: Langfuse (prompt/version registry & tracing), Helicone/Promptfoo/RAGAS/DeepEval for quality & cost metrics.
• Background & orchestration: Inngest/Trigger. dev or Temporal; platform crons (Cloudflare/Vercel/EventBridge/Cloud Scheduler).
• Observability: OpenTelemetry traces + Sentry/Datadog/Grafana Cloud logs/dashboards.
• Delivery: GitHub Actions CI/CD, per-PR previews; GrowthBook/Unleash feature flags/A-B tests.
• Auth & enterprise: Auth. js/Clerk/Stytch for auth; WorkOS/Okta for SSO/SAML when needed.
• Comms integrations: Resend/Postmark/Mailgun (email), Twilio/MessageBird (SMS), Google/Microsoft Graph (calendar).
• AI-accelerated dev: Cursor, Claude, Copilot/Cody—used deliberately with human-in-the-loop reviews.

Your responsibilities

• Own the architecture for a secure, multi-tenant SaaS, choosing a managed-first platform (e. g. , Vercel/Cloudflare/Fly. io/Render or AWS/GCP/Azure) that balances speed, reliability, and cost for a small team.
• Build end-to-end: TypeScript/React app, APIs, data models, background jobs, and real-time features.
• Ship AI capabilities: RAG search, entity dedupe/matching, summarization, recommendations—designed with evals, guardrails, and cost/latency budgets.
• Establish engineering standards: testing, code conventions, CI/CD, feature flags, observability, RBAC, audit trails, data privacy (consent/DNC, export/delete).
• Collaborate & lead: translate product goals into technical plans and help hire/mentor 3–6 engineers as we scale.
• Day 30 - Architecture doc + lightweight threat model with platform choice rationale; Monorepo scaffold (Turborepo/pnpm, Biome/lint) + CI (GitHub Actions); Auth/roles live; first UI flows (companies/contacts); seeded Postgres/Prisma; Baseline AI: embeddings pipeline + eval harness wired into Langfuse; initial traces and cost/latency dashboard.

About us:

We’re building a modern, AI-first CRM for recruiting. True greenfield with rapid iterations and real ownership. Goal: MVP in ~60 days, harden for scale by Day 90. You’ll be the cornerstone engineering hire working directly with the founder.

Why this role:

• Massive ownership: set the technical direction and standards from day one.
• Real product impact: ship AI features that recruiters use daily (search, summaries, matching, automations).
• Founding-level equity and a fast path to Head of Engineering / early CTO.

Compensation & equity

• Competitive salary + 5–10% founding-level equity
• Milestone bonuses tied to Day-60/90 deliverables.
• Change-of-control protection and refreshers at next financing.

How we work

• Small, senior, high-trust team; short cycles, small diffs, preview-before-merge.
• AI-accelerated dev, measured by quality, safety, and cost—not just speed.
• Security by default and data privacy from day one.

Process:

• Intro chat → Systems/product deep-dive → Practical pairing or time-boxed take-home (≤3 hours) → Founder loop → Offer.

Apply:

• Send GitHub/portfolio/LinkedIn and a short note on a 0→1 or LLM-powered feature you shipped (what changed and why).