Product Engineer

🇧🇷 🇦🇷 🇨🇴 🇲🇽 🇵🇾 🇨🇱 🇨🇷 🇵🇪 🇺🇾 Exclusive to candidates based in Brazil, Argentina, Colombia, Mexico, Paraguay, Chile, Costa Rica, Peru, and Uruguay

Requirements

Must-haves

- 5+ years of software development experience

- Experience owning at least one system in production

- Experience with React, Next.js

- Experience with Node.js

- Experience with Python

- Experience with Postgres

- Experience with AWS, including regulated or restricted environments requiring self-hosted services

- Experience building LLM applications (e.g., tool-using agents, retrieval and reranking, context management, long-running sessions)

- Experience evaluating AI systems (e.g., deterministic checks, LLM judges, threshold calibration)

- Experience with DevOps practices (e.g., CI/CD pipelines, containerization, infrastructure as code with Docker, Terraform, Pulumi, SST, etc.)

- Experience shipping products that real users rely on, with a clear account of post-launch outcomes

- Proficiency with AI-assisted development workflows (e.g., Claude Code) and informed views on their strengths and limits

- Ability to make UX and product calls on problems that arrive without a spec or a mockup

- Ability to tell a broken evaluation harness apart from an underperforming model

- Deep knowledge of production-grade system design and the failure modes that surface at scale

- Deep understanding of system-wide visibility (e.g., cloud services, security, service dependencies, observability)

- Strong communication skills in both spoken and written English

Nice-to-haves

- Startup experience

- Experience designing, building, and operating MCP servers

- Experience with regulated cloud environments (e.g. AWS GovCloud)

- Experience with data pipelines or ML infrastructure

- Experience with multi-tenant or whitelabeled products

- Background in workforce development or edtech platforms

- Bachelor's Degree in Computer Engineering, Computer Science, or equivalent

What you will work on

- Own the architecture of your domain end to end: data modeling, API design, cloud infrastructure, and CI/CD

- Build and release full-stack features spanning front-end, back-end, and infrastructure

- Develop production AI agents with sessions that stay open for weeks, per-turn context assembly, database-writing tool calls, persistent memory, and runtime redirection by non-engineers without code changes

- Expand the evaluation systems that track whether those agents improve or regress, keeping the harness trustworthy enough for the metrics to carry meaning

- Develop hybrid matching and search: parallel sparse and dense retrieval, reranking, recency handling, and looping real usage signals back into scoring

- Create the interfaces layered on top: streaming chat surfaces, admin and operator consoles, and several whitelabeled deployments delivered from one build

- Own the DevOps lifecycle: containerization, deployment pipelines, monitoring, and cloud environments, including one regulated setup where third-party APIs are off-limits

- Design, ship, and run MCP servers to integrate legacy systems and gradually retire them

- Define and enforce standards for code quality, security, performance, and tenant isolation in a SOC 2 Type 2 certified environment where certain deployments face stricter requirements

- Partner with product and leadership on build-vs-buy calls, technical risk, and architectural direction

- Stay connected to the people using the product and own whether the work actually landed

- Our Stack: TypeScript, Next.js, React, tRPC, Postgres with Drizzle, Tailwind and a shadcn-based design system, Python, AWS, Pulumi, SST, Docker, Qdrant, Bedrock, Turborepo, pnpm, Vitest, Playwright