Nikhil Phadke

-Led and managed an AI team across AI projects -Officially assigned mentor to AI team members -Enhanced a thermodynamic physics-informed optimisation algorithm using Pytorch -Developed RAG chatbot using Azure AI services, LangChain and postgreSQL/elasticsearch databases -Developed a forecast algorithm using Catboost/XGBoost to predict the production/consumption of solar energy with excellent accuracy. -Developed and finetuned Agentic AI applications using LLAMA/Claude-Sonnet

- Created an application feature to suggest the user a monthly optimum orientation of solar panels to yield maximum production based on just their location. - Worked on backend python development in Smart Energy Management

Current grade: 1,6 Master Thesis: Automatic Detection of Security Debt -Developed an active learning cybersecurity-focused NLP pipeline using PyTorch, transformers, CNNs, and LLM fine-tuning to detect security-related self-admitted technical debt. Research Project: Sentiment Analysis of Startup News - Applied NLP and data analysis techniques to classify sentiments in startup-related news articles. Project: Big Data - Web app for energy data processing; Apache Kafka/Flink, MySQL, Docker, GCP

Indian grade: 9,66 German grade: 1,2 Co-curricular: Team Captain/Brakes Domain Head - Designed and manufactured an electric all terrain vehicle from scratch to participate in the international intercollegiate motorsport competition SAE eBAJA. The team secured an international rank of 1 in a sub-event (Design Validation Event) and an overall rank 6 in SAE eBAJA ‘22 under my captaincy. - Mastered necessary skills of leadership, time management, negotiation, marketing