Pranava Kumar I

Pranava Kumar I

AI & Embedded Systems Engineer | Chennai, India

I build reliable engineering systems across aerospace, machine learning, quantum computing and data engineering — from flight computers and orbital prediction to AI applications and structured public-data pipelines.

What I Build Space Systems — Developed flight-ready avionics for the ROXα rocket; evaluated by ISRO / IN-SPACe / ASI, awarded "Excellent"; top 0.8% of 73,000+ participants. Physics-Informed ML — SMART-DEORBIT: PINN-based orbital decay prediction with under 1 km altitude error and ~100× speedup over DOP853. Quantum Computing — qkdpy: Python library for BB84, E91, CV-QKD and quantum network simulation; 200+ tests. PyPI. Applied AI — LLM-assisted anomaly detection across 76 ESA-AD telemetry channels and satellite-imagery rooftop solar detection with PicoLens. Data Engineering — Building Apify actors that transform fragmented public records into clean, typed JSON datasets for AI agents and analytics. Why Data Engineering?

Public data is often accessible but difficult to use reliably. Different agencies publish similar information through inconsistent formats, identifiers and classifications.

I focus on normalization: unified schemas, stable record identities, source provenance, deduplication, deterministic parsing and delta tracking.

My actors prioritize:

No LLMs in the data extraction path. Reliable, traceable and typed outputs. Isolated source failures. Change detection across scheduled runs.

My goal: Turn messy public records into dependable data infrastructure for developers, AI agents and analysts.

Connect

GitHub · LinkedIn · PyPI