Akash Bakshi User

Designed and developed an enterprise Agentic AI conversational platform for nationwide property search and customer assistance, achieving 30% lower latency and 15% productivity gains. Built persistent conversational memory with MySQL, tool-augmented agents for lead management/CRM workflows, and asynchronous low-latency request handling with FastAPI and Python.

Deployed an open-source LLM using an on-premises inference pipeline to reduce inference costs, and optimized the inference architecture to lower latency for image/HTML/text generation. Built a FastAPI backend for end-to-end request handling and dynamic prompt-based template rendering for production content delivery.

Built a web scraping pipeline with BeautifulSoup to extract and preprocess data into a searchable knowledge base. Implemented a document ingestion pipeline with intelligent chunking and semantic embeddings stored in Pinecone for similarity search, and optimized a low-latency LLM inference/template generation pipeline for dynamic content delivery.

Grade: CGPA: 8.10/10

Completed B.Tech in Computer Science and Engineering (2019–2023) with a CGPA of 8.10/10.

Grade: CGPA: 7.58/10

Completed XII (2018–2019) with a CGPA of 7.58/10 at J.K. Guru Gobind Singh School, Kunjwani.