Sujith Sudhakar

Motivated Cyber Security postgraduate with a passion for privacy-first AI systems and hands-on exposure in privacy-preserving ML. Eager to contribute to an organization where I can apply my knowledge in threat prevention, data security, and ethical hacking to protect digital infrastructure.

- TryHackMe – Cyber Defence Pathway

AI-Driven Privacy-Preserving Model for Children's Data Protection in Smart Learning Environments
Tools: Python, TensorFlow, PyTorch, Opacus, NumPy, Pandas
Dataset: Student Performance Dataset (Kaggle)

Description:
- Designed and implemented an AI framework integrating Differential Privacy to safeguard sensitive student data.
- Developed baseline (TensorFlow) and privacy-preserving (DP-SGD via PyTorch + Opacus) neural networks.
- Demonstrated effective privacy-utility trade-offs (ε ≤ 2.0) with minimal performance loss (MAE ~66).
- Visualized privacy budgets and model performance, aligning with GDPR and UNICEF data protection guidelines.
- Delivered a replicable, ethics-aligned AI pipeline applicable in EdTech and high-risk domains.

Key Learnings:
- Applied privacy-preserving ML techniques under real-world constraints.
- Gained experience in ethical AI, data protection regulations (COPPA, GDPR).

Date of Birth: 30 April 2003
Availability: Immediate
Willing to Relocate: Yes