Atul Singh

Problems worked on:

• Multi-agentic chat-bot for network management (US patent)
• QLoRA based training of Large language model for enterprise chat-bot
• Curated Alpaca dataset for model training on company's proprietary data
• RAG agent a enterprise LLM based Chat-bot and optimized it for faster response time.
• Machine learning based energy saving mechanism for enterprise building (US Patent)
• User location detection for in-building DAS Networks (US patent)
• Radio node clustering techniques to enhance coverage with minimal radio nodes (RNs) for in building solutions (US Patent)
• Implemented machine learning-based UE positioning, optimizing feature selection for precise location resolution.

• Analyzed and debugged NB-IoT systems with an in-depth understanding of physical, L2, L3, and core network layers. Created demo for customers using IoT devices which would transit data over our Radio access network.
• Led system integration and test automation for NB-IoT development from scratch to deployment, for European Space Agency.