Available for selective engagements

Edge AI, LiDAR & Digital Twins → Shipped

From concept to working systems across robotics, sensing, and cloud — with a bias for clean architecture, measurable outcomes, and shipping.

Start a project → See recent work → Let's talk! Email LinkedIn

Coming Soon

Field notes

The mindset

Practical notes on building with discipline—less theatre, more shipped outcomes.

operators > theorists focus beats vanity ship small, learn fast

What I do

Tightly scoped engagements or end‑to‑end builds. Always pragmatic, testable, and documented.

Advisory & Architecture

Clarify goals, pick the right stack, create a pragmatic roadmap.

DiscoveryRef ArchitectureDue Diligence

Prototype → Production

Rapid proof‑of‑concepts that scale cleanly to MVP and beyond.

POCMVPSRE/Observability

Edge–Cloud Pipelines

Reliable data flows with MQTT/DDS, ETL, and stream processing.

GoPythonAuth/OAuth2

Simulation & Digital Twins

Synthetic data for safety, speed, and smarter models.

Isaac SimCARLAScenario DSL

AI/ML & RAG Ops

Searchable knowledge and explainable answers over operational data.

LangChainVector DBGuardrails

DevSecOps & Compliance

Pipelines with gates that earn trust: tests, datasets, and audits.

CI/CDISO/SOC2SBOM

Recent work & experiments

A few glimpses from the last 12–18 months. Ask for deeper case studies and references.

Continuous GHG Monitoring (LiDAR)

Quantum gas LiDAR - Edge/Cloud stack for continuous CH₄ monitoring; dataset‑gated firmware releases and remote ops.

RM‑CW LiDAR • PLR/LTS • Edge(Core)→Cloud

Synthetic Factory + RAG Layer

Isaac Sim scenes generating RGB/LiDAR/telemetry; RAG answers ops questions with lineage.

AMRs • Fusion • Retrieval

MQTT → NinjaOne Bridge (Go)

Bridge heterogeneous sensor payloads to NinjaOne custom fields with OAuth token refresh and dynamic routing.

Go • MQTT • OAuth2

Smart‑City Traffic Digital Twin

3D city twin; edge analytics reduce bandwidth; cloud RAG explains congestion in plain English.

Omniverse/CARLA • Streaming

Streamlit + Supabase Fleet Insight

Commissioning vs live tables, queryable telemetry, clean API layer.

Supabase • FastAPI

Cybersecure Firmware & PLR

Defined PLR, ECR/ECO flow, gated track; reduced subcontractor dependency YoY.

Quality gates • Hardening

How we work together

Short cycles, measurable outcomes, clean hand-offs.

1. Diagnose

Understand constraints, success metrics, and risks. Create a short plan.
2. Prototype

Build a lean slice across sensing, perception and edge-cloud flow.
3. Prove

Benchmarks, vaidate data quality, and gated checks before scaling.
4. Scale

Harden, observe, and document. Transition pilot to sustained roadmap.

About me

Telesto Digital Director Best Digital Tech BBC — Prometheus Prometheus — Paper Pipeline Guild Award

Hi. I’m Puneet Chhabra — Board-aware software leader (ex-CTO) with 15+ years building and shipping AI-powered, real-time industrial platforms across utilities, O&G, and infrastructure. I build practical systems where sensing meets software: LiDAR, robotics, simulation, and edge-to-cloud analytics. I lead with clarity and ship with discipline. Passionate about leveraging technology for social impact and sustainability.

Former Head of Software at a deep, green-tech LiDAR company (GHG monitoring/PLR/ECR/ECO).
Former Higher Scientist at BAE Systems — Innovation award for my work on autonomous systems.
Led edge–cloud data pipelines for robotics fleets, including synthetic data gen and RAG layers.
Experienced advisor for startups on architecture, due diligence, and go-to-market strategies.
Hands-on with Go/Python, ROS2, MQTT/DDS, AWS, Supabase, Streamlit, FastAPI.

Double-digit ↓External contractor spend YoY

Cybersecure & LTSFirmware hardening & PLR

AMR & PerceptionEdge-AI, LiDAR & Synthetic loops

LiDAREdge AIDigital TwinsRAGDevSecOps

Tell me a bit about your project

I usually reply within one working day (UK time). Please include goals, constraints, and timeline.