Field Cto

Own the intelligence of the product. You're one of the first AI engineers in the room: the person who decides how our agent thinks, proves it with evals, and ships it to merchants the same week. About the role Yavendió builds Yago, an AI seller that closes deals on WhatsApp and Instagram 24/7, giving SMBs across LatAm and Brazil a seller that never sleeps.

The agent is not a feature bolted onto our product. The agent IS the product: if it stops selling, revenue stops. We're hiring a Founding AI Engineer to own agent quality end to end.

We're past the prototype phase: a Rust v2 agent platform in production, durable orchestration on Temporal, thousands of real conversations flowing through it. What we need now is the person who makes the agent measurably better at selling, week after week. This is a high-ownership, high-ambiguity role with a real growth path: as the team scales, this seat is designed to grow into technical leadership.

If you want tickets handed to you, this isn't it. What you'll do Own the agent core: prompts, tools, model routing, memory and conversation policy across our sales-agent runtime (Rust) and merchant copilot Build the eval loop we ship against: regression suites on real conversations, LLM judges, per-merchant quality scoring. Your bar is closed sales per conversation, not demo quality Ship across the stack when the problem demands it: Rust services, Python tool servers, our Next.Js dashboard Help land the v1 → v2 migration (Python/LangGraph to Rust + Temporal), then delete the old world Own reliability where it hurts: the WhatsApp/Instagram messaging pipeline, model regressions, edge-case behavior, incidents Work directly with founders and merchants to decide what the agent should do next, and why Help set the engineering bar, the culture, and the hiring standard as the team grows What we're looking for 5+ yrs of professional engineering on production systems, with real LLM products shipped: agents, tool use, RAG, evals.

Opinions formed by traffic, not demos Obsessive eye for detail. You're the person who catches the edge case in the diff, the token that leaks, the eval that's quietly lying Research-grade AI depth: you read the papers and model cards, you know why a technique works, not just that it does Always at the frontier of tooling. New models, new agent frameworks, new coding agents: you've tried them the week they shipped, and you use AI coding tools heavily every day Strong in Python or TypeScript today, excited to go deep in Rust (we'll teach you the codebase, not the curiosity) Proof of ownership beyond a job title: founded something, was early somewhere, maintains open source people depend on, or ran a system at real scale Fluent Spanish or Portuguese; working English Hardtechnical skills Python · TypeScript · Rust (or the drive to learn it fast) · PostgreSQL · Temporal or durable-execution patterns · LLM agents and tool use in production · Evals · Anthropic/OpenAI APIs · Docker · Git Nice to have WhatsApp ecosystem experience: Cloud API, BSPs, or the open-source stack around it Conversational commerce, social commerce, or B2B SaaS background You've hired or mentored engineers before Why this role The agent is the company.

Your work shows up in merchants' income the same week it ships A modern stack most AI engineers never touch: Rust agent runtime with typed guardrails, Temporal orchestration, frontier models in production An agent-native team: our own Claude Code plugin carries the team's conventions, an autonomous night watch triages production errors, PRs are reviewed by humans and agents Founding equity and a designed growth path into technical leadership Direct line to founders and customers. No layers, no bureaucracy