I’m an AI Engineer with 5+ years of experience building agentic automation and end-to-end LLM/RAG applications that help businesses move faster with less manual effort. I’ve worked across LLM-powered document intelligence, inference services, and evaluation/observability pipelines, leveraging Python, LangChain/LangGraph, and modern ML tooling to deliver measurable improvements in workflow throughput and response speed. In my recent work at Automation Anywhere, I design RAG systems and multi-tool AI agents that orchestrate enterprise documents, APIs, and business context while using guardrails and confidence-based human-in-the-loop escalation. Earlier, I built predictive models, recommendation systems, and computer vision/NLP pipelines, and I enjoy turning complex data problems into reliable production systems with strong evaluation and monitoring.

Yashaswi Rajesh Patki

I’m an AI Engineer with 5+ years of experience building agentic automation and end-to-end LLM/RAG applications that help businesses move faster with less manual effort. I’ve worked across LLM-powered document intelligence, inference services, and evaluation/observability pipelines, leveraging Python, LangChain/LangGraph, and modern ML tooling to deliver measurable improvements in workflow throughput and response speed. In my recent work at Automation Anywhere, I design RAG systems and multi-tool AI agents that orchestrate enterprise documents, APIs, and business context while using guardrails and confidence-based human-in-the-loop escalation. Earlier, I built predictive models, recommendation systems, and computer vision/NLP pipelines, and I enjoy turning complex data problems into reliable production systems with strong evaluation and monitoring.

Available to hire

I’m an AI Engineer with 5+ years of experience building agentic automation and end-to-end LLM/RAG applications that help businesses move faster with less manual effort. I’ve worked across LLM-powered document intelligence, inference services, and evaluation/observability pipelines, leveraging Python, LangChain/LangGraph, and modern ML tooling to deliver measurable improvements in workflow throughput and response speed.

In my recent work at Automation Anywhere, I design RAG systems and multi-tool AI agents that orchestrate enterprise documents, APIs, and business context while using guardrails and confidence-based human-in-the-loop escalation. Earlier, I built predictive models, recommendation systems, and computer vision/NLP pipelines, and I enjoy turning complex data problems into reliable production systems with strong evaluation and monitoring.

See more

Work Experience

AI Engineer at Automation Anywhere
July 1, 2025 - Present
Designed Python-based agentic process automation services integrating LLMs with Amazon Bedrock, AWS S3, REST APIs, and enterprise context to orchestrate 20K+ monthly AI-assisted workflow executions. Implemented LangChain RAG pipelines with workflow-aware retrieval, improving retrieval-grounded response accuracy by 18% in evaluated scenarios. Built LangGraph-based agent orchestration across bots, APIs, documents, and business applications to reduce manual intervention by 27%. Created document intelligence pipelines for LLM extraction/classification/validation, achieving 94%+ field-level extraction accuracy across 100K+ documents annually. Added self-healing automation that detects UI changes via application-state and visual signals and switches execution paths to reduce failures by 22%, with human-in-the-loop controls, confidence scoring, and guardrails monitored via LangSmith. Refined prompts, retrieval strategies, and execution flows using Ragas and LangSmith to reduce AI response lat
Machine Learning Engineer at Infinite Infolab
August 1, 2019 - August 1, 2023
Developed predictive ML models in Python using scikit-learn, XGBoost, and pandas, evaluating 10+ classification and regression approaches across customer, operational, and transactional datasets. Built computer vision pipelines with OpenCV and CNNs for image classification and anomaly detection, processing 50K+ images and reducing false positives by 18% through model tuning. Implemented NLP solutions using NLTK, spaCy, and BERT for text classification and entity extraction across 100K+ support and business text records. Designed data preparation and feature engineering workflows using pandas/NumPy/SQL, transforming 20+ attributes into reusable ML features and reducing feature-generation time by 22%. Built recommendation models using collaborative filtering and gradient-boosting techniques, evaluating multiple approaches with Precision@K and offline validation. Deployed ML inference services through Docker and REST APIs, integrating models into 5+ applications with typical sub-300ms inf

Education

Master of Science in Artificial Intelligence at Yeshiva University
August 1, 2023 - May 1, 2025
Bachelor in Information Technology at Amravati University
August 1, 2016 - November 1, 2020
Master of Science in Artificial Intelligence at Yeshiva University
August 1, 2023 - May 1, 2025
Bachelor in Information Technology at Amravati University
August 1, 2016 - November 1, 2020
Master of Science in Artificial Intelligence at Yeshiva University
August 1, 2023 - May 1, 2025
Bachelor in Information Technology at Amravati University
August 1, 2016 - November 1, 2020
Master of Science in Artificial Intelligence at Yeshiva University
August 1, 2023 - May 1, 2025
Bachelor in Information Technology at Amravati University
August 1, 2016 - November 1, 2020

Qualifications

Add your qualifications or awards here.

Industry Experience

Software & Internet, Professional Services, Education, Computers & Electronics