About
Built by someone who has sat on the other side of the security review.
I am Johnson Kosgei, a data and AI leader with more than 15 years in enterprises. I help regulated mid-sized organisations define their AI strategy, turn it into a roadmap and operating model, and move AI from pilot to production by doing the identity, data-access and evidence work their reviewers ask for.
I have spent more than 15 years building data, analytics and AI capabilities in pharma and healthcare, entertainment and technology, first in the United States and now in Austria.
Most recently I led data, analytics and AI for a pharmaceutical and nutraceutical group: Copilot and data agents on Azure, measurement of adoption and business impact, and responsible-AI training. I also worked on the roadmap, architecture and governance behind those efforts. Good pilots stalled. The model had not failed. Identity, permissions and review evidence were not ready, and nobody yet owned the operating model that would keep the work going.
That gap is where most of the value of AI is lost, and it is work I know from the inside. It closes only when strategy, operating model and delivery are designed together, so I work on all three. I do it for regulated mid-sized companies, at a fixed price.
Read-only access for assessments, plain-language reports, no lock-in and no hourly billing. Every script, template and report I produce is yours.
Executive and data leaders who need a strategy their teams can run, security and risk leads at regulated companies, and Microsoft partners who need a specialist for the governance part of a rollout.
Track record
More than 15 years of making data and AI work inside real organisations.
- From 2025Data, Analytics and AI Manager, a pharmaceutical and nutraceutical group. Integrated 60+ data sources into Microsoft Fabric, deployed Copilot and data agents on Azure, and ran AI enablement and training.
- 2022 – 2025Data and AI Practice Lead and co-founder, Northrift Solutions. Data science, governance and master-data projects for retail and healthcare clients, and a retrieval-based AI service for farmers that works without internet access.
- 2021 – 2023Head of Product, Data and Engineering, xy.ai. AI and machine-learning pipelines over satellite imagery and health records at scale.
- 2017 – 2021Director, Data Engineering and Analytics, SAG-AFTRA Plans. Led a team of 14 and built a governed healthcare data warehouse covering more than a million participants.
- 2009 – 2017Manager, Data Analytics, Red Bull. Santa Monica and Salzburg. A portfolio of 30+ analytics projects and a global analytics community.
Recognition. Ventana Research Leadership Award (2015). Esri Special Achievement in GIS Award (2013). Keynote speaker, Esri User Conference, Salzburg (2017).
What I hold myself to
Five rules, and five things I do not do.
- Evidence over assurance. I show the record, not my opinion of it.
- Least access needed. Read-only wherever the work allows.
- Your data stays yours. So does everything I build for you.
- Plain language. In every report.
- Handover by design. Your own people own the result.
- I do not bill by the hour.
- I do not resell licences or models.
- I do not lock you in.
- I do not promise a regulatory outcome.
- I do not take over your accountability.