Role Description
We are building a desktop application for single-cell genomics analysis and visualization, with built-in AI agents and skills. We are adopting an open-source AI coding agent’s architecture as our foundation.
You will join the backend team to build this product on top of our existing GPU-accelerated analytics stack. The role spans multiple languages and layers — from the desktop app’s server logic down to the analytics engine — and you will help shape the architecture as the product grows.
Single-cell biology knowledge is not required. We welcome engineers at any seniority who show strong ownership and system design intuition.
Key Responsibilities
- Build the desktop app backend, adopting and extending our forked open-source architecture.
- Design and implement local MCP servers and Skills that expose our analytics capabilities to in-app and external agents.
- Bridge the new desktop layer with our existing analytics services — handling subprocess lifecycle, streaming results, and error propagation.
- Collaborate with the team to grow the product’s features and capabilities as needs evolve.
- Write clear design docs and PR descriptions; review code; investigate and adopt new tools as the product evolves.
Requirements
Must have
- Strong backend engineering skills in at least two of: TypeScript, Python, or another mainstream backend language.
- Comfortable reading and forking large codebases; able to extend an existing system without rewriting it.
- Experience with inter-process communication, subprocess management, and HTTP API design.
- Hands-on experience with at least one of: MCP/agent tool systems, desktop apps, or local-first sync engines.
- Strong system design intuition — you can break a feature into parts, name the boundaries, and explain the trade-offs.
- A habit of investigating before asking — you read the code, run the tests, and form a hypothesis before requesting help.
Nice to have
- Experience with single-cell genomics, ontologies, or bioinformatics tools.
- Experience with GPU computing or large-scale scientific data.
- Contributions to open-source projects.
- Exposure to code tracing, sandboxing, or safe LLM-code execution.
Personal qualities we value
- Hardworking and reliable.
- Curious and open-minded.
- Strong problem solver.
- Communicates proactively.
- Thinks about long-term maintainability before writing code
