Sarat Kumar Manugula
Offering over 8+ years of experience in Information technology and software development field focusing on and Machine Learning, Deep Learning ,NLP and Gen AI in Mechanical Domain, Medical, BFSI
- Role
AI Solutions Architect
- Years of Experience
9.3 years
Skillsets
- PowerBI
- Jira
- Keras
- LangGraph
- LlamaIndex
- LLMs
- LSTM
- Microservices
- MLOps
- NoSQL
- Oop
- Pinecone
- HuggingFace
- Prophet
- PySpark
- Python
- rag
- react
- REST
- XgBoost
- YAML
- Yolov8
- Azureiot
- C#
- AWS - 1 Years
- Azure - 2 Years
- OpenCV
- pandas
- PyTorch
- SQL
- TensorFlow
- .NET
- Autoencoders
- Azuredevops
- Azurestreamanalytics
- MLFlow - 2 Years
- CI/CD
- crewAI
- Deepsort
- Docker
- Faker
- FastAPI
- Flask
- GCP
- Github
- Githubactions
Professional Summary
9.3Years
- Oct, 2024 - Present1 yr 6 months
AI Solutions Architect
Ntt Data
Applications & Tools Known
Azure
AWS
Power BI
Mlflow
Huggingface
Work History
9.3Years
AI Solutions Architect
Ntt DataOct, 2024 - Present1 yr 6 months
Achievements
- Stood 3rd in SAS Hackathon
- Pegasus Award Jan 2016
Major Projects
12Projects
Citi Bank Credit Card Recommendation using Agentic AI
Architected multi-agent AI recommendation system analyzing customer spending patterns; improved conversion by 45% and reduced latency to 400ms. Tech Stack: LangGraph, Azure OpenAI GPT-4, FastAPI, PostgreSQL, Docker.
PNB RFP Risk & Fraud Detection Platform
Designed high-level architecture for 10 AI/ML banking use cases including assumptions, risks, dependencies, deviations, and resource loading; ensured RBI compliance. Tech Stack: Power BI, Draw.io, PPT.
MCP for Reviews Multi-Platform Review Analytics
Built MCP-based system aggregating reviews from Amazon, eBay, Walmart, and Google using GPT-4o with dynamic tool orchestration. Tech Stack: MCP 2.0, Python, GPT-4o, Marketplace APIs.
Synthetic Data Imputation Privacy Preserving Data Generation
Designed LLM + Faker based synthetic data platform from minimal seed data with validation agents, fallback logic, and relational integrity. Tech Stack: Python, Groq/OpenAI, Faker, Pandas, FastAPI, Docker.
Invoice Validation Intelligent Document Processing
Built AI-powered invoice validation system using OCR and document AI, reducing manual effort by 80%. Tech Stack: Azure Document Intelligence, GPT-4o, Python, FastAPI, SQL, Azure.
Energy Forecasting Predictive Analytics Platform
Architected multi-horizon forecasting platform achieving MAPE < 5% across UK and Spain energy datasets. Tech Stack: XGBoost, Prophet, LSTM, FastAPI, AKS, Docker.
IGO/NIGO Agent Insurance Claim Processing
Automated insurance claim validation using GPT-4o and document intelligence, reducing review time by 70%. Tech Stack: Azure OpenAI, Azure Computer Vision, FastAPI, React, YAML.
Advanced RAG Framework Test Case Generation
Designed hybrid RAG system generating automated test cases with 70% effort reduction. Tech Stack: LlamaIndex, GPT-4, Pinecone, FastAPI, Docker.
Battery Life Forecasting IoT Predictive Analytics
Built federated learning-based battery health prediction system achieving 92% accuracy. Tech Stack: SARIMAX, XGBoost, Prophet, TensorFlow Federated, Azure IoT Hub.
IAQ Analytics Indoor Air Quality Intelligence
Developed real-time IAQ analytics platform achieving 25% energy savings using ML and physics-based models. Tech Stack: Azure Stream Analytics, Azure IoT Hub, Python, Power BI.
Asset Analytics Cognitive Predictive Maintenance
Designed federated anomaly detection framework reducing unplanned downtime by 40%. Tech Stack: TensorFlow Federated, Autoencoders, Azure IoT Edge, Docker.
Video-Based Queue Analysis Retail Computer Vision
Implemented real-time CV system using YOLOv8 and DeepSORT reducing customer wait time by 30%. Tech Stack: YOLOv8, OpenCV, DeepSORT, NVIDIA Jetson, Docker.
Education
Bachelor of Electronics and Communication Engineering
Aditya Institute of Technology and Management (2013)
Certifications
Azure dp100 certification