Peeku and FuelSaathi are now the primary work: private desk wellness and evidence-led car-ownership tools. The earlier intelligence systems and writing remain part of the practice.
What I do
I build products around daily frictions and difficult systems.
The current work is Peeku, a private desk-wellness coach for Mac and Windows, and FuelSaathi, source-linked E20 and car-ownership tools for Indian drivers.
The wider build record includes Sanjaya, Sanket, AI explainers, and essays on configuration and public decision intelligence.
Why it is different
The same discipline applies at different scales.
Past work around senior ministerial, PSU, and institutional communication environments taught me that evidence, hierarchy, voice, timing, and risk matter. The public intelligence products grew from that context.
Peeku and FuelSaathi apply the same discipline to everyday use: make a bounded problem easier to understand and act on without overclaiming what the product knows.
2010Google
Selected into Google, an early technology-company chapter that remains part of the background to the work.
April 2026Sanjaya + Sanket
Public energy and cyber-risk intelligence systems established the first visible product layer.
July 2026Peeku + FuelSaathi
Independent products for desk wellness and Indian car ownership became the current focus.
NowBuild, learn, improve
Keep the products useful, preserve their privacy and evidence boundaries, and let real use shape the next iteration.
The angle
This is an independent product practice, not a generic AI newsletter. The work begins with a concrete user problem and follows it through to a usable surface: healthier desk habits, a better answer about a car, or a clearer read of a complex system.
The working hypothesis: useful software earns trust by making its boundaries clear. Peeku is explicit about local processing. FuelSaathi is explicit about source-linked vehicle guidance. Sanjaya and Sanket remain public examples of the same commitment to a legible, inspectable surface.