Defence tech · Experiment
SENTINEL
Tactical intelligence synthesizer
Awaiting responseSubmitted to the iDEX SPARK grant · Ministry of Defence, India
Capabilities
Offline-first
Runs on local Ollama. No cloud needed. Works in field conditions with no internet.
Threat classification
Auto-categorizes threats: aerial, cyber, supply chain, IED, border, NBC, reconnaissance.
Confidence scoring
Every assessment includes a confidence score. No black boxes — you know what the AI knows.
Multi-source synthesis
Ingests raw field reports, intel briefings, and sensor data. Synthesizes into actionable assessment.
Report generation
Auto-generates formatted reports with IDs, timestamps, and recommended actions.
No data leaves the device
All processing happens locally. Sensitive intel never touches external servers.
Status
Research (done)
Studied defence AI gaps, threat assessment workflows, and offline ML deployment
Build (done)
Built the full system — classifier, assessment engine, report generator, UI
Submit to iDEX (done)
Official submission to iDEX SPARK Grant under Ministry of Defence, India
Awaiting Response (active)
Submitted. Waiting to hear back from the Ministry of Defence.
Field Deployment (pending)
If approved — pilot deployment with field units
Simulated demo
A UI preview that runs entirely in your browser. It sorts your text with simple keyword rules and fills in a report template — it is not the real model. The full system runs offline on a local model via Ollama, or on Novita AI.
Simulated output for demonstration only — not a real threat assessment. SENTINEL has been officially submitted to iDEX and is awaiting a response from the Ministry of Defence.