Agentic Ai Application Engineer

Name: Manish Y

Email: **************@*****.***

Phone: 414-***-****

LinkedIn: www.linkedin.com/in/manish-yamsani

Professional Summary

AI/ML Engineer with 4+ years of experience designing, developing, and deploying enterprise-scale Generative and Agentic AI solutions across healthcare and finance. Hands-on with RAG pipelines, LLM orchestration, and multi-agent frameworks (LangChain, CrewAI, AutoGPT). Skilled in MLOps (MLflow, Databricks, Jenkins, GitHub Actions) and cloud platforms (Azure ML, AWS SageMaker, GCP Vertex AI, Vortex AI). Also, with AI/ML Engineer with strong expertise in Python, SQL, and MLOps, in transforming data into actionable insights through analytics, experimentation, and automation.” Hands-on experience with advanced AI frameworks such as Lang Chain, Semantic Kernel, Crew AI, and AutoGPT. Adept at building NLP pipelines, predictive models, and intelligent copilots, with deployments that meet compliance standards (HIPAA, SOX, FINRA, GDPR). Recognized for translating business challenges into AI-driven products, optimizing workflows, and collaborating with cross-functional teams to deliver measurable impact. Technical Skills

• Programming: C#, Python, SQL

• ML & AI Frameworks: TensorFlow, PyTorch, Scikit-learn, Hugging Face Transformers, XGBoost, LightGBM

• Generative AI: Large Language Models (LLMs), Lang Chain, Retrieval-Augmented Generation (RAG), Prompt Engineering, Multi-Agent Systems, Crew AI, AutoGPT, Semantic Kernel, Copilot Studio

• Data Processing: pandas, NumPy, spaCy, NLTK, OpenCV

• MLOps & Deployment: MLflow, Airflow, Fast API, Flask, Docker, Kubernetes, GitHub Actions, CI/CD, Databricks, Kafka Streams

• Cloud Platforms: AWS (SageMaker, EC2, Lambda, Glue, S3), Azure Machine Learning, Azure Power Platform, GCP

• Databases: PostgreSQL, MongoDB, Redis, SQL

• Visualization & BI: Power BI, Plotly, matplotlib, seaborn Professional Experience

Client: Elevance Health Gen AI Engineer Jun 2023 – Present

• Designed and deployed LLM-based applications using Lang Chain, Crew AI, and Semantic Kernel to support healthcare workflows and enterprise automation.

• Developed AI solutions with LLM-based applications using LangChain, AutoGPT, Semantic Kernel, and CrewAl to automate decision-making and retrieval workflows, streamlining healthcare operations

• Built Retrieval-Augmented Generation (RAG) pipelines in Python with FAISS and Pinecone, enabling secure document retrieval and summarization.

• Built and deployed production-grade RAG and LLM applications with API integration layers using FastAPI and LangChain for real-time inference and automation.

• Fine-tuned open-source LLMs (LLaMA, Falcon, Mistral) for clinical and operations use cases, improving contextual accuracy under HIPAA compliance.

• Engineered multi-agent orchestration using AutoGPT and react-based patterns for autonomous claims and HR process automation, resulting in improved efficiency and reduced manual intervention.

• Built an evaluation setup with test datasets using LLM to track accuracy, errors, and speed of GenAI pipelines.

• Created Copilot-style assistants with Microsoft Copilot Studio, integrating with Power Automate and Power Apps to streamline HR, onboarding, and support workflows.

• Applied prompt engineering and chain-of-thought reasoning to reduce hallucinations and improve chatbot reliability.

• Deployed and monitored models in Azure ML and Vortex AI environments, optimizing cost and performance under enterprise governance.

• Deployed containerized models with Docker and Kubernetes, using GitHub Actions for CI/CD. Key Achievements:

• Reduced claims processing time by 35% through multi-agent automation.

• Improved clinical text summarization accuracy by 28% using fine-tuned LLMs.

• Cut chatbot error rates by 40%, boosting customer satisfaction scores. Client: State Street AI/ML Engineer Sep 2021 – May 2023

• Developed ML models for fraud detection and risk analytics, ensuring alignment with SOX and FINRA compliance.

• Built NLP pipelines for contract classification and sentiment analysis using Hugging Face Transformers and spaCy.

• Automated retraining workflows with Python, MLflow, AWS Lambda, and S3 versioning to handle data drift.

• Deployed RESTful APIs using Flask and FastAPI, supporting secure, scalable model access for enterprise users.

• Deployed APIs with Flask and Fast API on AWS SageMaker and EC2, integrating models into large-scale financial applications.

• Built RAG evaluation framework for retrieval accuracy and hallucination control.

• Applied explainable AI methods (SHAP, LIME) to provide transparency in audit reviews.

• Monitored production models using AWS CloudWatch and built anomaly alerts.

• Designed feature pipelines with Python, Pandas, NumPy, and SQL to process high-volume datasets. Key Achievements:

• Detected and prevented $2M+ in fraudulent transactions annually.

• Reduced model drift issues by 25% through automated retraining pipelines.

• Accelerated contract analysis workflows by 30% with NLP automation. Client: SLK Software ML Engineer Jun 2020 – Jul 2021

• Built predictive ML models using Scikit-learn and XGBoost for client projects in finance and retail.

• Developed and deployed APIs with Flask, containerized with Docker, and orchestrated using Kubernetes.

• Supported MLOps practices including CI/CD pipelines, model versioning, and monitoring with AWS SageMaker.

• Processed structured and unstructured data using pandas, NumPy, and SQL.

• Delivered reports and dashboards using Power BI to track business KPIs. Key Achievements:

• Improved demand forecasting accuracy by 18% for retail client.

• Enhanced API deployment efficiency by 20% using containerization.

• Cut reporting time by 25% through automated Power BI dashboards. Education

• Master of Computer Science – Concordia University-Wisconsin, USA (2024)

• Bachelor of Technology in Computer Science – MLR Institute of Technology, India (2021)

Contact this candidate