Pratham Singla

AI Engineer and Data Scientist with around 2 years of production-focused experience (+1 year experience of freelance) at an AI-native company, specializing in LLM-powered systems, autonomous and multi-agent AI, RAG pipelines, and voice AI at scale. Shipped 5+ products to production — including a live autonomous calling agent (5,000+ calls) and an enterprise multimodal GPT serving 300+ daily active users — working across the full ML lifecycle, from data engineering through LLMOps, evaluation, and AWS deployment.

AI Engineer and Data Scientist at UMENIT Solutions LLP (2024-11 – Present)

• Shipped 5+ AI products to production end-to-end within a 15-member cross-functional team, owning work from architecture design through deployment, monitoring, and iteration.
• Designed and deployed LLM-powered automation systems for unstructured data processing, agentic workflows, and real-time business insights using advanced RAG architectures.
• Optimized performance and cost across production voice and LLM systems — cut voice latency, LLM/token spend, and RingEmAll's per-minute call cost, and stabilized throughput under API rate limits and server load with a custom rate limiter; maximized LLM output quality through prompt engineering and LLM-based classification for more accurate intent routing.
• Led LLMOps and MLOps — prompt optimization, RAG tuning, evaluation (RAGAS, DeepEval), and experiment tracking (MLflow); built REST APIs with FastAPI and Flask, deployed on AWS (EC2, Lambda, S3) with Docker.

Bachelor of Engineering in Computer Science – Chandigarh University (2020-09 – 2024-06)