Key Responsibilities :

- Lead deep technical research across AI security areas including model behaviour analysis, provenance, and unlearning.

- Define research hypotheses, experimental design, and validation criteria for differentiated AI security techniques.

- Drive development of proofs of concept, benchmarks, and experimental frameworks.

- Act as subject-matter expert for model fingerprinting, anomaly detection, derivation analysis, and unlearning verification.

- Review and guide work of applied researchers and research engineers to maintain scientific rigour.

- Contribute to system architecture discussions in collaboration with the CTO / Chief Scientist.

- Author and co-author technical papers, research reports, and intellectual property.

- Stay current with global research trends and continuously revalidate internal research direction.

Required Qualifications & Experience :

1. Education :

- PhD (CS, AI/ML, Crypto, Cybersecurity) from Premier Institute with strong research credentials in AI/ML, Computer Science, Cybersecurity, or related fields.

2. Experience :

- 8 to 12+ years of experience in AI/ML research or applied security research.

- Proven experience leading research initiatives or acting as technical lead for complex research programs.

- Earlier Experience: R&D labs, Research projects, Product companies, AI Projects, Enterprise solutions, AI Engineering.

- Publications in top tier conferences (mandatory).

Required Technical Skills :

- Strong expertise in large language models, fine-tuning, LoRA/adapters, and retrieval-augmented generation.

- Deep understanding of model internals including weight-space analysis, activation patterns, and interpretability.

- Hands-on experience with AI security techniques such as fingerprinting, backdoor detection, model similarity, and provenance analysis.

- Knowledge of model unlearning methods, verification techniques, and privacy assurance trade-offs.

- Proficiency with ML experimentation frameworks (e.g., PyTorch/JAX) and reproducible research workflows.