AI product consulting
Investigate the real need, test feasibility, define a pilot, and make the tradeoffs explicit. We can also provide focused technical product leadership while a team finds its footing.
Explore product workWe turn messy problems and complex information into AI products and decisions people can test, understand, and use.
We join at the point where product judgment, language technology, and evidence need to meet.
Investigate the real need, test feasibility, define a pilot, and make the tradeoffs explicit. We can also provide focused technical product leadership while a team finds its footing.
Explore product workFind failure patterns, build useful test sets, compare approaches, and decide when human review matters.
Explore evaluationTurn technical questions into studies, workshops, and courses that help teams make better decisions.
Explore research and learningDifferent settings, same discipline: understand the context, make the system testable, and expose its limits.
Multilingual extraction and evaluation work in a production clinical NLP platform, with careful attention to performance, regression coverage, and release decisions.
Read the case Civic technology · Independent nonprofitFact-checking support, manipulation detection, and narrative analysis through Kate's work with Aletheos.
Read the case Search · Dutch-language documentsSemantic search, query expansion, and text quality work from Kate's time at Faktion.
Read the caseSignal & Sense is a new practice. The cases describe Kate's earlier roles and nonprofit collaboration.
Each engagement is scoped around a decision you need to make. The sequence adapts; the evidence stays visible.
Name the users, stakes, constraints, and failure modes.
Build a bounded prototype, research plan, or evaluation path.
Check it against relevant data and the people doing the work.
Show the tradeoffs and agree what changes next.
Signal & Sense is led by Ekaterina (Kate) Loginova, an NLP engineer and technical product lead. Kate works between research, engineering, product teams, and the people a system is meant to help.
The practice is deliberately focused. Larger implementations may involve separately agreed partners; you always know who is doing the work.
Meet Kate