AI · government architecture · energy security · public intelligence

AI systems for India's decision rooms.

I build AI-native intelligence surfaces for the places where policy, energy, executive communication, and public systems meet. The work here is not demo AI; it is a public proof layer for institutional decision-making.

Inside the system Ministerial and PSU context

Worked around policy, executive communication, and energy-sector decision environments where language, evidence, and hierarchy matter.

Built the systems Sanjaya + Sanket

Public intelligence surfaces for Brent, Indian pricing, energy infrastructure, and cyber risk around the oil-and-gas system.

Writes the layer Policy, diplomacy, technology

Long-form arguments, executive drafting, explainers, and product copy from the same operating vantage point.

Thesis

The serious India AI opportunity is decision intelligence.

India does not need another thin chatbot demo. It needs AI surfaces that can sit inside files, briefs, dashboards, sector reviews, crisis maps, cabinet-facing summaries, and operating workflows. That is the line connecting this site: energy intelligence, cyber-risk intelligence, AI explainers, and writing for institutions that make consequential decisions.

Flagship system

Sanjaya

The flagship public intelligence surface for Indian oil and gas: pricing, crude, infrastructure, companies, and policy, seen whole.

Open live surface →
Cyber-risk system

Sanket

Executive cyber-risk intelligence, India energy surface map, public signal tracker, and browser posture scan.

Open dashboard →
Core essay

India AI is a decision-intelligence problem

The argument behind the work: AI becomes valuable when it makes public and institutional systems more legible.

Read thesis →
India AI

AI governance and public decision intelligence

AI for India-scale institutions: government, energy, policy, workflows, and shipped systems.

Oil and gas

Sanjaya + Sanket

Energy intelligence, cyber-risk intelligence, pricing, infrastructure, and public oil-and-gas signals.

AI compute

Chips, stacks, inference, and post-training

Start with the full AI compute stack, then drill into chips, terms, bottlenecks, and training pipelines.

MCP tools

Claude-connected operating surfaces

Small tools that connect models to budgets, routines, files, and real personal context.

Recent writing

Built AI-native public intelligence systems across energy, cyber risk, finance, AI compute, and government-facing decision workflows. 3 products · 5 essays · 5 explainers · 1 briefing

Start anywhere

Elsewhere