Bhagyalaxmi Yogendra Sawantdesai
Applied AI Engineer | Generative & Agentic AI Innovator | LLMs, RAG, Predictive Analytics, Responsible AI
- Role
- Senior Analyst Data Science and Machine Learning(Genai & Agentic Ai Engineer) at Infosys
- Location
- Mumbai, MH, IN
- LinkedIn followers
- 500 followers
About Bhagyalaxmi Yogendra Sawantdesai
Generative & Agentic AI Specialist Driving the future of AI — from Generative AI and Agentic AI to Data Science, Deep Learning, and NLP. I design intelligent systems that transform industries and deliver measurable impact.What began as a fascination with robotics has evolved into a mission: building AI that serves human needs with clarity, responsibility, and innovation. Over 6+ years, I’ve delivered solutions across healthcare, retail, and financial services — architecting multi‑agent systems, RAG pipelines, LLM platforms, predictive models, and automation frameworks that redefine how organizations leverage AI Key Achievements- Enterprise Workflow BottlenecksArchitected multi‑agent orchestration frameworks with LangChain, LangGraph, CrewAI, and AutoGen to automate complex decision flows.→ Result: 85% workflow automation, freeing teams from repetitive tasks and enabling faster enterprise operations- Fragmented Knowledge SystemsDesigned retrieval‑augmented generation (RAG) pipelines with hybrid vector search (ChromaDB, Pinecone, FAISS) and fine‑tuned LLMs (RLHF, DPO, ReAct).→ Result: 90% query accuracy, reducing knowledge silos and accelerating insights across departments- Scaling AI from Pilot to ProductionDelivered end‑to‑end cloud‑native MLOps pipelines (AWS Bedrock, Azure ML, GCP Vertex AI) with Kubernetes, MLflow, and FastAPI.→ Result: Seamless deployment of AI models from experimentation to production, ensuring explainability, scalability, and sustainability- Risk & Compliance in Regulated DomainsEngineered anomaly detection, defect prediction, and bias‑mitigation frameworks with LangSmith/LangFuse observability.→ Result: 30% reduction in defect leakage, embedding Responsible AI practices into enterprise pipelines- Human‑AI Collaboration GapsBuilt autonomous agent ecosystems that integrate with RPA, APIs, and enterprise data lakes.→ Result: 40% faster delivery cycles, enabling hybrid teams (human + AI agents) to deliver measurable ROI.
Experience
Senior Analyst Data Science and Machine Learning(Genai & Agentic Ai Engineer)
Aug 2022 — Present · Hyderabad, IN
LLM-Powered QA Engine: Developed GPT-4 powered test synthesis platform using few-shot learning and prompt chaining;implemented auto-evolving prompts and context windowing, reducing QA effort by 75% while expanding coverage to 90% • Multi-Agent AI Systems: Architected autonomous healthcare workflows using LangChain, AutoGen, and CrewAI; implemented ReAct agents with Chain-of-Thought reasoning and LangGraph orchestration for complex pharmacy operations; achieved 85% automation in clinical decision support. • Advanced RAG BigQuery Integration: Engineered domain-specific RAG system integrating GCP BigQuery (Cloud SDK) with hybrid vector search (ChromaDB); implemented automated schema discovery and SQL validation pipeline with semantic routing; achieved 90% query accuracy and built comprehensive regression suite for continuous data validation across healthcare knowledge bases; reduced query development time by 60% through automated schema introspection. • Intelligent Automation: Built Figma UI automation pipeline combining computer vision (OpenCV/OCR) with LLM reasoning; integrated accessibility testing through Playwright with MCP routing; reduced sprint cycles by 40% through automated code generation. • Intelligent No-code Low-code Automation: Built low-code RPA Agent Builder platform using Flowise for autonomous workflow automation; implemented intelligent execution engine with OCR/NLP for data processing, contextual decision-making, and self-correction capabilities; enabled continuous learning through automated feedback loops. • Deep Learning,Ensemble Methods & Optimization: Built solution automating data standardization and transformation for claims and provider data, improving accuracy and throughput; implemented transformer-based deep learning models (BERT) with multi-output classification for unstructured data insights, applying reinforcement learning and Bayesian optimization to enhance model accuracy.Also developed ML models for anomaly detection.
Education
Aditya Polytechnic Ratnagiri
Diploma, Electronics and Telecommunication
2013 — 2016
Great Lakes Institute of Management
Post Graduation Program, Data Science
2019 — 2020
St. Francis Institute Of Technology
Bachelor of Technology - BTech, Electronics and Telecommunication
2016 — 2019
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.