Azhar is a home for focused software experiments—exploring how large-scale public data, thoughtful interfaces, and applied intelligence can make everyday systems calmer and more capable.
Self-contained public data indexes, astronomical timekeeping utilities, and applied-intelligence prototypes in active exploration.
Public discovery index for 50M+ registered Indian enterprises.
A zero-database, statically indexed directory built over open government registration datasets—making millions of micro, small, and medium enterprises searchable by district, sector, and activity code.
Astronomical Hijri calendar, solar prayer times, and interactive astrolabe dial.
A zero-database astronomical timekeeping engine computing Fatimid (Misri) Hijri dates, proportional solar hours, and solar noon (zawaal) across global coordinates with web and offline mobile parity.
AI-powered residential rental discovery and deterministic shortlisting funnel.
A multi-stage property triage and evaluation engine combining discrete search filters with hybrid inference—pruning hard deal-breakers, scoring interior photos and soft preferences, and guiding household shortlists.
Every experiment is engineered as a self-contained instrument—built to stay in orbit with minimal moving parts.
Every experiment starts with one sharp problem.
We favour focused tools that do one job well over sprawling, high-maintenance platforms.
Pre-compute, statically index, and deploy to the edge.
Where possible, we eliminate runtime databases and moving parts—keeping infrastructure simple, fast, and durable.
Models and agents work best behind the scenes.
We use intelligence to structure messy public data, evaluate imagery, or accelerate tedious workflows without getting in the user’s way.