About Sedes

Sedes is an AI-native property operations platform, cross-asset by design for real estate operators in

Europe and abroad.

Tenants talk to us entirely over WhatsApp: an AI agent handles maintenance requests, scheduling, and questions end to end, escalating to human operators when needed. Our goal is to create a Digital-Twin of all properties within a portfolio, so our Agents together with operators may orchestrate everything: rentals, obligations, compliance, cleaning and maintenance planning, and tenant communication. Live in production with paying clients; we are a Meta WhatsApp Tech Provider. The role

Our current technical lead is transitioning out and will run a full structured handover with you. You will inherit a working, revenue-generating product and own it completely: architecture, roadmap, delivery,

and eventually the engineering team. This is a true #1 technical seat. Depending on fit, this role becomes CTO. What you'll own

- A Next.JS 16 / React 19 dashboard (TypeScript, TanStack Query, Zustand, Tailwind v4) on Vercel
- A serverless API: 14 Express-on-Lambda microservices (Node 22, TypeScript) behind API Gateway,

Cognito auth, EventBridge scheduling, all on AWS eu-south-1

- Supabase Postgres with PGVector RAG; no ORM, typed data layer
- The AI core: an OpenAI Agents SDK orchestrator with ~35 tools and 4 personas, multi-provider model routing (OpenAI + Gemini via Vertex), eval harnesses with LLM judges
- WhatsApp Cloud API integration: templates, webhooks, encrypted Flows
- A separate document-AI platform (Bun/TypeScript, CDK, Qdrant, MinIO, ERPNext) that extracts structured financial data from ingested documents. What you'll do
- Ship product weekly across the full stack; this is a hands-on coding role




- Own reliability of a system real tenants depend on daily
- Mature the engineering foundation: CI/CD, E2E testing, and deploy automation currently don't exist,

and we'll be honest that this is your first 90 days

- Drive the AI agent roadmap: new tools, personas, eval coverage, model routing
- Work directly with founders and clients to turn operator pain into features. You Must have
- Senior TypeScript, full stack (our entire codebase is TS)
- AWS serverless in production (Lambda, API Gateway, Cognito, IAM)
- Shipped LLM-powered features to real users: agents with tools, RAG, evals. This is core product,

not a bolt-on

- Postgres fluency
- High agency: you have run systems where you were the last line of defence. Premio points
- WhatsApp Cloud API / Meta platform experience
- Multi-tenant SaaS design
- CDK / Docker / EC2 ops, ERP integrations (ERPNext/Frappe)
- Italian language or Italy market context
- Comfort with AI-assisted development workflows (we build with Claude Code daily)
- Prior founding or first-engineer experience Open problems you'd tackle
- Zero-to-one CI/CD and E2E testing across three repos

2.

Agent reliability: eval coverage and regression gating for a 35-tool orchestrator

- Replacing scheduled-sweep patterns with proper queueing as volume grows
- Scaling multi-tenancy from a handful of clients to dozens. Interview process
- Intro call with the founder (30 min)
- Technical deep-dive with the outgoing tech lead: walk through our real architecture, no leetcode

(60–90 min)

3.

Paid trial project: ship one small real feature in the codebase (1–3 days)

4.

Offer

How to apply

Send a short note with: something real you've shipped with LLMs (link or writeup), your GitHub or portfolio, and one thing in our stack you'd change first. No cover letters.

📌 Technical Lead (Roma)
🏢 SEDES
📍 Roma