We're looking for someone who doesn't just review a sprint plan but can sit down and build the feature themselves if the team gets stuck. Someone who measures their week by what was shipped, not by what was discussed. You care less about being the person who designs the system and more about being the person who gets a team of engineers to build it well, on time, without the codebase quietly rotting underneath them.
The architecture for our agentic, GenAI-powered platform already exists; the principal architect owns those calls. Your job is to turn that architecture into working software: breaking it into a plan a team can execute sprint over sprint, writing full-stack code yourself when it's faster than explaining it, catching the failure modes and shortcuts in code review before they become production incidents, and being the person who can honestly tell the product owner whether a deadline is real or wishful thinking. Success looks like a team that ships predictably, not because the plan was perfect, but because you kept it honest.
Responsibilities:
- Break down architecture and product requirements into a sprint-level plan your team can actually execute and be accountable for hitting it.
- Write production code yourself across the full stack, frontend and backend, when the team needs it, not just direct others who do.
- Run code review as a real quality gate: catch bugs, shortcuts, and design drift before they ship, not after.
- Unblock engineers on day-to-day technical blockers, unclear requirements, or conflicting priorities so the team keeps moving.
- Report delivery status, risks, and tradeoffs honestly to the product owner, including when a timeline needs to change.
- Mentor engineers on the team through code review and pairing, raising the team's baseline over time.
Requirements:
- 6+ years of engineering experience, including time leading or mentoring a team.
- Demonstrated hands-on full-stack engineering, able to personally build and debug across frontend and backend, not just architect on a whiteboard.
- A track record of leading or mentoring a team of engineers through shipping production features on a schedule, with good quality.
- Ability to translate an architecture or UX design doc into a concrete, sequenced delivery plan.
- Strong code writing and review instincts, able to spot failure modes, security, performance, reliability issues, and shortcuts.
- Leverages linting, unit testing, and API testing to take ownership of code and design quality.
- Comfortably owning day-to-day tradeoffs between speed, quality, and scope, and communicating those tradeoffs clearly to non-technical stakeholders.
Tech Stack:
- Backend: Python (FastAPI, async/await, Pydantic), PostgreSQL, REST API, MongoDB, Redis, pytest, Kafka/Google Pub/Sub, OpenAPI/Swagger.
- Frontend: Next.js, React, TypeScript, JavaScript, TanStack Query, HTML5/CSS3 Jest, SSR/SSG/CSR/ISR patterns, frontend performance, and bundle optimization.
- Infra and Tooling: Docker, CI/CD pipelines, Git-based code review workflows, SonarQube, Prometheus and structured logging, Kubernetes (GKE), and GCP (Cloud SQL, GCS, and GKE).
- AI Systems: Working knowledge of RAG pipelines and multi-agent orchestration patterns (to execute against, not design), vector search and embeddings, and message-queue/event-driven patterns.
Good to Have / Bonus Points:
- Experience working directly against an architecture defined by someone else, rather than owning architecture end-to-end.
- Exposure to agentic/GenAI or RAG-based systems in production.
- Experience in a fast-moving, early-stage engineering org where the process is still being built.

