Founding Engineer

About the role

Bloom is the brand layer for agents. It turns a brand’s identity, guidance, and assets into shared context that agents and products can use. That context evolves with the brand, alongside the tools where it is created and managed.

We’re a three-person, YC-backed team building this infrastructure from the ground up.

The Role

We're looking for an engineering generalist with depth in backend and AI systems. You'll help decide what to build, shape the architecture, and own what you ship in production. The immediate work centers on agents, data, and infrastructure, with room to follow a problem wherever it leads across engineering.

This is a full-time, in-person role in San Francisco. We're building our core team here. Visa support is available.

You'll work on problems across:

- Agents. Build agents, tools, memory, and workflows that create, maintain, and apply brand systems.

- Multimodal data. Build ingestion and source-understanding systems for websites, social media, and files, working across text and visual material.

- Storage and retrieval. Develop the data models, indexing, search, and retrieval that make brand context useful over time.

- AI quality. Build evaluations and feedback loops that help us understand quality, investigate failures, and improve our systems.

- Developer interfaces. Make Bloom useful to other products and agents through our API and MCP.

- Production infrastructure. Improve security, latency, and reliability. Build frontend surfaces and internal tools when they help solve the problem.

You'll make decisions about what we build now, what we simplify, and what can wait.

What we're looking for

- You have strong fundamentals and production depth. You've built and operated backend and AI systems that people depend on. You can explain what you personally designed, why you made those choices, what broke, and how you fixed it.

- You understand how the parts of a system fit together, reduce scope and complexity, and build primitives that remain useful as the product grows. You can ship a small, correct first version without losing sight of what it needs to become.

- You have product and customer judgment. You understand who or what will use what you build, question whether it solves the right problem, and are comfortable talking directly with customers to improve the product.

- AI tools are a substantial part of how you build. You still understand the code, scrutinize the result, and take responsibility for its correctness and maintainability.

- You question requirements and explain your reasoning. You're comfortable with direct technical discussion, defending a decision, and changing your mind when the evidence is better. You care more about getting it right than being right.

- You move with urgency and follow through. You make progress through ambiguity, surface problems early, and stay with difficult work through production.