Our work
Systems we’ve built, and how they work.
AI agents, assistants and automations we designed, built and tested, from voice front desks to private inbox routers.
Showing 8 projects
Also built
Smaller builds and experiments.
Self-checking research agent team
Four LLM agents clarify a question, search the web, score their own confidence and re-search when evidence is thin.
Agent operations console
A tool-using agent that searches, reads documents and writes files live, with token and cost tracking per run.
AI career copilot
A five-step pipeline that turns a CV and a job ad into a match score, skill gaps and a tailored cover letter.
Resume-to-job fit analyser
Schema-validated analysis with prompt-injection and no-fabrication guardrails, and editable tailored drafts.
Research-paper integrity checker
Seven checkers flag impossible statistics, recomputed p-values and fake or retracted citations (hackathon build).
Property price estimator
End-to-end machine learning: scraped listings, a gradient-boosting model, an API and a dashboard.
Have a process like these in mind?
Book a free call. We’ll look at how your business runs and tell you honestly what’s worth automating, and what isn’t.