• We are developing advanced code classification models using DL and statistical techniques, significantly improving accuracy.• We enhanced classifier training with code stylometry techniques, optimizing through iterative testing and evaluation.• Collaborated with the research team to implement innovative AI-driven solutions, Generative AI and NLP.

• I researched and reported results around the potential of LLMs for malware analysis/detection from a de-compiled code.• My role was to conduct a literature survey, design scenarios, and implement the experiment workflow using Python.• Aimed to investigate how good new LLM (GPT3.5/4) tools are in the position of an expert malware analyzer.

• I engaged in R&D for voice synthesis, focusing on modelling and fine-tuning ML models to align with product needs.• Deployed scalable ML pipelines on AWS SageMaker and integrated LLM solutions with data workflow• Partnered in implementing a RAG system, leveraging LLM benefits on the company’s database for better decision-making.

• Collaborated with a team of five to develop SigmaPilot, a Generative AI code reviewer as a VSCode extension, optimizingcode analysis using top-performing NLP model Vicuna-13B.• Deploy Vinuca-13B model, prompt engineering, and leveraging my knowledge in data privacy and security.• Integrated the GenAI chatbot (Vinuca-13B model), and design scenarios to enhance the code review process on the extension,maintaining high standards of code privacy by leveraging local infrastructure.

Functional testing of trading application "Adatra," as well as security testing and security solutions.Designing more than 250 test cases and test scenarios via the Microsoft Azure platform.Database security solutions such as encryption, access control, etc.

Producing educational material on Java-Python OOP design patterns in the blog format.