SDE 3

About Libra AI

Libra AI is an early-stage, venture funded startup building an AI Superagent for work- semantic search and agentic workflows across 250+ integrations to streamline work across teams like product, HR, finance, sales, and operations. Our team is rapidly growing and committed to shaping the future of intelligent workplace automation.

The role

You'll be heads-down writing code on the hardest parts of the stack while also shaping how the rest of the engineering team thinks, builds, and ships. You'll architect and own the backend systems that power agentic workflows at scale (orchestration pipelines, retrieval layers, eval infra) and raise the bar for everyone around you.

What we're looking for

Must-haves

• 4 to 5+ years in a strong SDE role, shipping production systems
• Hands-on experience with AI orchestration frameworks (AI SDK, LangChain, LlamaIndex, or equivalent)
• Solid grasp of vector databases, RAG pipelines, embeddings, chunking strategies, and retrieval evals
• Core AI fundamentals: quantization, attention, tokenization, inference economics, fine-tuning vs prompting trade-offs
• Strong JavaScript/TypeScript chops; MERN stack is a prerequisite
• Architected scalable backend systems including async pipelines, queues, caching, and db design
• Has worked on systems that have seen real scale (traffic, data, or concurrency)
• Strong communicator who can articulate technical trade-offs clearly and bring a team along
• Demonstrated leadership: has mentored engineers, owned architecture decisions, or led a pod
• Genuine passion for AI. You read papers on weekends, ship side projects, and your GitHub looks like a green forest. Bonus if those side projects are AI-flavored.

Bonus

• DevOps fluency (Docker, K8s, CI/CD, infra-as-code)
• Prior experience at an AI-first company or shipping LLM products
• Exposure to graph databases, agent frameworks, or multi-model routing

What we offer

• Top-of-market cash plus meaningful early-stage equity
• A real seat at the table. You'll architect and lead, not just implement
• Directly work with the founder and the core team on problems that actually matter
• High-agency environment with the resources of a well-funded pre-seed company