01 What is AXL
Axiom Labs (AXL) is an independent computational research identity — a one-person research collective operating at the intersection of pure mathematics, algorithms, and applied systems.
The name reflects a core principle: build from axioms. Every system should have a rigorous theoretical foundation before implementation. No shortcuts, no duct tape.
AXL is the identity under which I compete in national hackathons, develop independent research projects, and publish technical work. It allows me to maintain a coherent research identity across institutions, competitions, and collaborators.
Currently focused on: Gaussian Process theory and optimal transport, graph neural networks for fraud detection, survival modeling for credit risk, energy systems analytics, and machine learning with mathematical guarantees.
02 Research Pillars
Designing efficient algorithms with provable guarantees. Not just "it works" but "it works because..."
Convex and non-convex optimization for real-world systems — energy, infrastructure, logistics.
ML with mathematical foundations — not black-box models but systems you can reason about.
Computational tools for energy monitoring, disaggregation, and policy optimization.
03 Trophy Cabinet
04 AXL Projects
Non-Intrusive Load Monitoring framework for appliance-level energy disaggregation. Random Forest anomaly detection + LSTM load prediction. Smart plug with MQTT + FastAPI backend.
Verification-first framework for post-collection waste accountability using QR-linked batch proofs, geotagging, and timestamped logs to close accountability gaps in waste management.
Optical sensing-based system for appliance-level energy inference. Modeled consumption behavior using time-series patterns and cost analytics. Competed solo — Ranked 5th out of 1000+ participants.
OBD-II telemetry pipelines for hybrid efficiency analysis. Rule-based hybrid mode switching strategies to maximize energy recovery and minimize fuel consumption.
Real-time detection and classification of solar X-ray flares fusing ISRO XSM/Chandrayaan-2, ESA PROBA-2/LYRA, and NASA GOES XRS data. Rolling-minimum background estimation, Savitzky-Golay smoothing, Bayesian confidence scoring, and a calibrated 16-feature Random Forest classifier.
Competed in the EcoCraft Policy Challenge at IIT Kanpur Policy Conclave 2026. Reached Round 3 with proposals at the intersection of environmental policy and computational analysis.
AXL — Axiom Labs
Every system built under AXL starts from axioms, not assumptions.