Products we've built. Outcomes that matter.
We don't just ship features. We help founders build things investors fund and users stay for.
DecoverAI
AI Search / LegalThe Problem
Legal teams were spending hours manually searching thousands of documents for relevant case precedents. Keyword search returned too much noise and missed conceptually related content.
What We Built
A semantic search platform using LLMs and vector embeddings (PgVector) that lets users query documents in plain English. The system ingests, chunks, and indexes legal documents, then surfaces contextually relevant results ranked by semantic similarity.
Outcome
The product helped the client raise $2M in seed funding shortly after MVP launch. Now in active use with law-domain clients.
Mappie AI
AI Productivity / SaaSThe Problem
Product teams lose days translating stakeholder briefs into structured, engineering-ready tickets. The handoff between product and dev was a constant source of delay and misunderstanding.
What We Built
An AI-powered project management tool that takes raw requirement inputs — PDFs, notes, meeting transcripts — and generates structured user stories, acceptance criteria, and task breakdowns ready for sprint planning.
Outcome
Early users report 60–70% reduction in requirements grooming time. Bootstrapped and growing.
Vectro AI
Sales Intelligence / LLMThe Problem
Sales managers had recordings of every call but no scalable way to extract patterns, objections, or coaching insights.
What We Built
An AI pipeline that ingests sales recordings, transcribes them, and runs structured LLM analysis — extracting objection patterns, sentiment shifts, talk ratios, and deal risk signals surfaced in a clean dashboard.
Outcome
Sales teams now get coaching insights across their entire pipeline in the time it previously took to review one call.