SAVANTH GATTU

Prompt Engineered and fine-tuned Large Language Models (LLMs) for Lineaje AI and SBOM360 products, focusing on optimizing vulnerability detection in software supply chain security. Developed, trained, and fine-tuned LLM models, including Mixtral 7B, BERT, Llama 13B, and T5 utilizing PEFT and Quantization Techniques for converting natural language to MongoDB queries. Designed and implemented a semantic search engine by integrating AI-based caching and fine-tuned LLMs. Designed and implemented APIs using FastAPI.

Developed a novel AutoML technique and utilized an ensemble of machine learning models to predict the most suitable employee for a given task in the ARP-BOT project. Established a system capable of automatically predicting the optimal resource from a pool based on various factors and historical data. Created deep learning architectures, including autoencoders and GRUs, for deinterleaving and classifying radar pulses. Developed an AI-based control unit (AICU) to proactively classify and mitigate threats.

Implemented a ROS-based mobile robot prototype with 2D LiDAR, SLAM, motion planning, navigation, and reinforcement learning for autonomous operation, complemented by a specialized ROS Android app enabling smartphone control.

Developed a website enabling worldwide remote management and monitoring of robot parameters using Internet of Robotic Things (IoRT), establishing seamless communication between the CC3220SF Wi-Fi module and an Angular website through WebSockets with MQTT Protocol.