We are a venture studio, which means we build companies rather than joining them. We currently have four active portfolio companies at different stages: a legal document drafting tool at $800k ARR, a customer support automation platform at Series A, a clinical trial protocol generator at seed stage, and a B2B sales intelligence tool that launched six weeks ago. All four are LLM-native. All four have NLP challenges that exceed what their current teams can solve independently. We hire a small number of NLP specialists at the studio level who work across the portfolio — spending three to six months at a company, solving the hardest NLP problem on their current roadmap, and then rotating to the next. The arrangement is unusual. It suits people who get bored fast, who like variety, who want to accumulate breadth across applications rather than depth in one product, and who can work with a different founding team every few months and earn their trust quickly. It doesn't suit everyone. If you've spent four years on the same NLP pipeline and want to go deeper, we are probably not the right environment. If you've been thinking that the most interesting NLP problems are spread across too many different domains to commit to just one, we might be exactly right.
Responsibilities
Solve the highest-priority NLP problem in each portfolio company during your rotation — scoped jointly with the founding team at the start of each engagement
Leave each rotation with the codebase, documentation, and institutional knowledge transferred so the resident team can maintain it
Contribute to shared NLP infrastructure at the studio level — embedding pipelines, evaluation frameworks, and fine-tuning tooling used across the portfolio
Participate in technical due diligence on new companies the studio is considering building or backing
Mentor junior engineers at portfolio companies during your rotation
Requirements
4+ years of NLP engineering across at least two different application domains — variety in your background is genuinely valued here
Hugging Face Transformers for fine-tuning on domain-specific tasks — legal, clinical, and enterprise data each behave differently
Python at a professional standard — you write code that another engineer can maintain after you rotate away
RAG pipeline design for document-heavy applications — you understand why different retrieval strategies matter for different document types
LangChain or LlamaIndex for building multi-step agentic and retrieval workflows
SQL for understanding how NLP outputs are stored and used downstream in each product
Comfort onboarding quickly into a new codebase and team every three to six months
Benefits
Work across four different NLP application domains in your first two years — legal, clinical, sales, and customer support
Full remote with optional travel to portfolio company offices for onboarding weeks (covered)
$105,000 – $128,000 base salary at the studio level
Small equity stake in each portfolio company worked on — vesting model explained in detail during interviews