I'm Charudatta Korde, a PhD researcher focused on deploying intelligent, secure AI systems on real hardware.
My work sits at the intersection of:
- 🧠 Machine Learning Systems
- ⚡ FPGA-based Acceleration
- 🔐 Cybersecurity & Trusted Computing
🎓 PhD Scholar @ NIT Goa
🛠 Open-source contributor & tooling architect
👨🏫 Mentor in applied AI, security, and systems engineering
I specialize in taking neural networks from notebooks to silicon, ensuring they are efficient, verifiable, and secure.
- PyTorch, TensorFlow, Keras
- Model optimization & deployment
- OCR, vision, and sequence models
- NNGEN-based neural synthesis
- HDL workflows (Verilog integration)
- Hardware-aware model design
- Malware analysis (Wireshark, Cuckoo)
- Secure automation pipelines
- Threat modeling & sandboxing
- Python, FastAPI, Flask
- Redis, Docker, GitHub Actions
- PowerShell, LaTeX, Lua
- 🧠 FPGA-Optimized Neural Network Architectures
- 🔗 Blockchain-enabled Edge Computing
- 🔐 Secure & Trusted AI Pipelines
- 📚 AI-assisted instructional design
🔗 https://github.com/charudatta10/task-runner-SDLC
Problem: Manual SDLC tracking causes delays and inconsistencies
Solution: Automated workflow orchestration system
Impact:
- ⏱ Reduced manual overhead by 45%
- 📈 Improved sprint predictability
Tech: Python, Flask, SQLite
🔗 https://github.com/charudatta10/devanagari-handwriting-recognizer
Goal: Accurate OCR for handwritten Devanagari scripts
Results:
- 📊 94% accuracy on real-world samples
- 🧠 Robust to writing variability
Tech: TensorFlow, Keras, NumPy
- ✨ VS Code for AI-Enhanced Note-Taking
Integrating snippets, workflows, and semantic indexing for research productivity.
- 📜 Mentored 6 postgraduate theses (2025)
- 🏃 Completed RG Marathon 5K (2024)
I'm open to:
- 🤝 AI × FPGA research collaborations
- 🧪 Applied ML systems projects
- 🔐 Secure AI & edge computing work
- 📚 Open-source & educational initiatives
SPDX-License-Identifier: LicenseRef-Charudatta10-Personal-Use-Copyleft-1.0
Personal Use only.
No commercial, institutional, or professional use without a paid license.
Strong copyleft — derivatives must retain identical terms and attribution.
© Charudatta Gurudas Korde
https://github.com/charudatta10