Sk Asaif Ali

AI/ML Engineer with 2 years of hands-on experience building and deploying agentic AI, RAG, and automation systems. Work spans retrieval design, LLM orchestration, evidence validation, APIs, and observability, with systems shipped from model selection through deployment, including solutions deployed in client environments.

Software Engineer, AI/ML at Bebo Technologies Pvt Ltd (2024-11 – 2026-08)

• Cut candidate screening-to-interview turnaround by 75% for a pipeline handling 100+ applications per opening by building the LLM-based candidate scoring and resume-parsing module for an autonomous recruitment agent, developed with a 3-person pod.
• Improved retrieval precision by an estimated 28% on an enterprise Agentic RAG system indexing over 10,000 documents by implementing hybrid dense/sparse search with Qdrant and cross-encoder reranking.
• Cut irrelevant retrievals by around 19% across roughly 5,000 monthly queries on the same system by implementing query rewriting and contextual retrieval before the LLM reasoning stage.
• Reduced manual legacy-code migration effort by an estimated 34% per module across a 200,000+ line codebase by contributing to the code-transformation and validation stages of a multi-agent migration system built with a 4-person team.
• Reached roughly 91% wound-boundary detection accuracy by fine-tuning a Masked Autoencoder (MAE) model in PyTorch on a curated dataset of 2,000+ wound images as part of a small clinical-AI team.

M.S. in Data Science – VIT AP University (2022-09 – 2024-10)