┌────────────────────────────────────────────────────────────────────┐
│ $ cat operator.log │
├────────────────────────────────────────────────────────────────────┤
│ │
│ handle → TensorNaut │
│ name → Tushar Jagatap │
│ location → Raver, Maharashtra, India │
│ role → AI / ML Engineer · GenAI Developer │
│ focus → Agentic AI · RAG Systems · Deep Learning │
│ mission → Build AI that doesn't just predict — it decides. │
│ status → 🟢 ONLINE · 💠Open to AI / ML opportunities │
│ goal → Ship production-grade, self-reasoning AI systems │
│ │
└────────────────────────────────────────────────────────────────────┘
|
AST-based code chunking, ChromaDB vector store, SentenceTransformers + Groq LLM. Chat with any repo.
|
TrueContext is a Retrieval-Augmented Generation (RAG) system that answers questions strictly from user-uploaded documents.
|
|
→ TensorNaut/Health-Monitoring-System Wearable AI — ESP32 sensors, 200K+ records, stress & heart disease prediction at 92.8% accuracy.
|
CNN-Transformer-CrossAttention on RAVDESS. 68.26% WA · 66.23% UA · Streamlit demo.
|
builds = {
"ExpenX" : "AI expense agent — auto-categorize, track, analyze",
"SER_System" : "Speech Emotion Recog — CNN-Transformer-CrossAttn",
"BDH_Paper" : "Sparse LLM ablation — Baby Dragon Hatchling arch",
} |
interests = [
"Agentic AI → ReAct loops, multi-agent systems",
"RAG → hybrid retrieval, reranking, eval",
"DL Arch → Transformers, Cross-Attention",
] |
