Articles, systems notes, and technical essays.
A flat writing index for market notes, engineering details, and first-principles arguments.
30 years of AT&C losses, told as one long engineering bug report
India has spent three decades trying to reduce power distribution losses. The number moved, but the deeper bug stayed: the sector still lacks a trusted record at the meter.
Canonical bytes: why JSON betrays you when you try to hash it
A signature only means something if every verifier can reproduce the exact bytes that were signed. JSON makes that harder than it looks.
The Engineering Discipline Autonomous UAVs Need Before Autonomy
Most autonomy startups want to start with mission AI. Valeon has to start one layer lower: requirements, sizing, energy margins, airframe constraints, and contested-link behavior that keep Chimera honest.
India needs 1.32 million chargers by 2030. The grid doesn't know yet.
India is at 29,000 public chargers. The 2030 target is 1.32 million. Everyone is debating where the boxes go. The harder problem is whether the grid can prove what they did.
What nobody tells you about the nanosatellite electronics market
I led flight-facing CubeSat electronics work at TSC Technologies between ages 21 and 23. Here's what that job actually looks like, what the Indian nanosatellite market doesn't say out loud, and which credentials carry weight versus which ones are resume padding.
The Systems Thesis Under My Companies
The same pattern keeps appearing across JouleBridge, Axiotelic, Valeon, Ferrite, Vishwa, ZeroOTP, Amber, and Parallax: useful systems produce evidence, expose boundaries, and can say no.
What Atlas Teaches Robotics Startups
Boston Dynamics' electric Atlas is a product lesson disguised as a humanoid reveal. The lessons that matter are not the flips. They are actuator commonality, thermal discipline, safety geometry, perception packaging, power operations, and the refusal to confuse demos with deployment.
Why JouleBridge now: smart meters, EV charging, and AI agents converging
Three infrastructure curves are moving faster than the evidence layer underneath them: 250M smart meters under RDSS, 1.32M EV chargers by 2030, and a $58B AI-energy-ops market. JouleBridge is the layer that makes the three coherent.