I'm Charudatta Korde, a PhD researcher focused on deploying intelligent, secure AI systems on real hardware.
My work sits at the intersection of:
- π§ Machine Learning Systems
- β‘ FPGA-based Acceleration
- π Cybersecurity & Trusted Computing
π PhD Scholar @ NIT Goa
π Open-source contributor & tooling architect
π¨βπ« 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
- π§ FPGA-Optimized Neural Network Architectures
- π Blockchain-enabled Edge Computing
- π Secure & Trusted AI Pipelines
- π AI-assisted instructional design
π https://github.com/charudatta10/task-runner-SDLC
Problem: Manual SDLC tracking causes delays and inconsistencies
Solution: Automated workflow orchestration system
Impact:
- β± Reduced manual overhead by 45%
- π Improved sprint predictability
Tech: Python, Flask, SQLite
π https://github.com/charudatta10/devanagari-handwriting-recognizer
Goal: Accurate OCR for handwritten Devanagari scripts
Results:
- π 94% accuracy on real-world samples
- π§ Robust to writing variability
Tech: TensorFlow, Keras, NumPy
- β¨ VS Code for AI-Enhanced Note-Taking
Integrating snippets, workflows, and semantic indexing for research productivity.
- π Mentored 6 postgraduate theses (2025)
- π Completed RG Marathon 5K (2024)
I'm open to:
- π€ AI Γ FPGA research collaborations
- π§ͺ Applied ML systems projects
- π Secure AI & edge computing work
- π 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 β derivatives must retain identical terms and attribution.
Β© Charudatta Gurudas Korde
https://github.com/charudatta10


