AI InternFlask REST services, thermodynamic feature engineering, and pytest validation for a compressor CBM platform.
00 / Origin
Shaunak
Rane
Machine-learning systems engineer
I build validated ML systems that turn complex physical and operational data into reliable decisions across industrial diagnostics, agent orchestration, graph learning, and distributed energy.
01 / Agent systems
Full-stack buildMulti-agentFastAPI
Aegis
investigation system
A full-stack claim investigation workflow with deterministic ingestion, asynchronous evidence and verdict stages, specialist market and news agents, and a reviewable dashboard.
Claim lifecycle2 stages
Discovery loop15 min
Specialist logicScout + Trending
The core product is the traceable state machine around the model: deduplicate, investigate, persist, review, and expose failure.
02 / Energy systems
Team projectRL controlDePIN
Gridium
protocol
A four-runtime microgrid prototype connecting a 15-node physics simulation, continuous-control DDPG agent, realtime 3D command center, EVM energy AMM, and Groth16 surplus circuit.
500msphysics tick rate
7D → 1Dstate to fee action
15simulated grid nodes
The prototype makes its simulation boundary explicit: the controller, market, proof path, and operator UI are integrated system components.
03 / Industrial intelligence
AI internshipIndustrial MLFlask
Compressor
evidence system
A condition-monitoring platform that separates what the thermodynamics support from what a polished forecast merely suggests.
3plant-specific contracts
3,879DFI rows reproduced
17/17validation gates
Forecasting remains an experimental visualization and never drives maintenance alerts.
04 / Graph intelligence
Independent studyGraph MLPhysics benchmark
TopoFlow
pore network
A permeability study where the topology is learned, the classical baseline stays visible, and pore heterogeneity decides which model deserves trust.
1,231pore-network graphs
5rock families tested
46.2%best MSE reduction
Benchmark result: learned topology helps in heterogeneous media; it is not a universal replacement for physics.
05 / Trajectory
What I bring
to a team
I move comfortably between investigation and implementation: data forensics, ML evaluation, APIs, Docker, responsive interfaces, and the documentation that keeps claims honest.
Software InternFastAPI backend, BOM pipeline, and Random Forest attendance prediction for food operations.
Independent workTopoFlow GNN, Aegis, and Gridium across backend, ML, and distributed-system design.
