Auris Lab builds and researches AI products for cybersecurity and business administration. You will take ideas from prototype to something people use every day, and bring the same practice to customers rolling out enterprise AI.
What you'll do
- Prototype and ship AI features: agents, retrieval over business documents and tool use with guardrails.
- Evaluate models and prompts properly, with test sets, not impressions.
- Design how AI agents get and give up access to systems, safely and on the record.
- Help customers deploy enterprise AI assistants and local LLMs, and measure whether they help.
- Write up what we learn for Auris Lab's research notes.
What you bring
- Five or more years in software engineering, with at least two building with LLMs in production.
- Strong Python or TypeScript, and experience with LLM APIs, embeddings and vector search.
- Good judgement on security and privacy when AI touches business data.
- You ship: small, working increments over long plans.
Nice to have
- Experience running open-weight models on your own infrastructure.
- Background in security tooling or business software (finance, HR, operations).
- Public work: open-source projects, papers or a technical blog.