Kamera Vamshi

AI/GenAI Engineer building distributed LLM pipelines, Agentic workflows, and cloud-based AI automation systems. Optimized LLM batch processing latency by 64% and reduced token inference costs by 35%+ across 150K+ document workloads through Dask-based parallelism, prompt engineering, and context optimization. Engineered RAG architectures, leveraged Vector Databases, and deployed scalable Machine Learning solutions using FastAPI and GCP.

AI / Software Developer at Tudip Technologies (2024-09 – Present)

• Engineered a head-tail partitioning strategy with Dask on GCP, reducing LLM API integration latency for high-volume document pipelines by 67%.
• Optimized LLM batch processing for 150K+ documents, slashing Google Reviews processing time by over 82 minutes per workload.
• Decreased Gemini API call count by 80% (from 6,667 to 1,306 calls) and reduced token consumption by 30% via automated batch consolidation and prompt engineering.
• Designed an Adaptive Clustering Strategy transitioning from KMeans to MiniBatchKMeans, achieving an 80% reduction in clustering time for million-record workloads.
• Developed and deployed production-grade RESTful APIs using FastAPI, handling high-throughput ingestion of 5,000+ requests/minute for AI automation workflows.
• Launched automated data retrieval via Google Cloud Functions, enabling event-driven LLM orchestration and increasing system throughput by 40%.

Data Analyst Intern at My Analytics School (2023-05 – 2023-07)

• Analyzed 10,000+ community acquisition data points to identify 3 critical drop-off points in the lead funnel, formulating targeted strategies that increased qualified lead inflow by 20%+.

Bachelor of Technology (B.Tech) – National Institute of Technology, Rourkela (2020-08 – 2024-05)