Machine Learning Generative Ai

Poojitha Vadde

# ***********@*****.*** +1-980-***-**** ð Poojitha Vadde

Professional Summary

Experienced Machine Learning & Generative AI Engineer with 4+ years of expertise in building, deploying, and scaling LLM-powered applications and end-to-end ML workflows. Specialized in NLP, Retrieval-Augmented Generation (RAG), and multi-agent orchestration using tools like LangChain, LangGraph, and Hugging Face. Proficient in MLOps, scalable deployment with Docker/Kubernetes, and CI/CD integration across AWS, Azure, and GCP.

Technical Skills

AI & NLP: GPT-4, GPT-3.5, Claude 3.5, Mistral, LLaMA 3, T5, BERT, GPT-2, ChatGPT, RNN, CNN, Transformers, Embedding Models, Natural Language Processing (NLP), Generative AI (GenAI), Large Language Models (LLMs), Retrieval-Augmented Generation (RAG), Prompt Engineering (Zero-shot, Few-shot, Chain-of- Thought), Knowledge-based Retrieval Systems, Deep Learning, Neural Networks Frameworks & Libraries: LangChain, LangGraph, LlamaIndex, Hugging Face, Autogen, SpaCy, NLTK, Rasa, Dialogflow, Streamlit, TensorFlow Keras, PyTorch, Pandas, NumPy RAG & Vector DBs: Pinecone, FAISS, RecursiveTextSplitter, Hybrid Retrieval APIs & Integration: FastAPI, Flask, Node.js, OpenAI API, Azure OpenAI API, REST APIs, Postman MLOps & Deployment: Docker, Kubernetes, MLflow, DVC, GitHub Actions, Jenkins, CI/CD Cloud Platforms: Google Cloud Platform (Vertex AI), Amazon Web Services (SageMaker, Lambda), Microsoft Azure (OpenAI, Blob Storage)

Languages: Python, SQL, Bash, JavaScript, TypeScript Other Tools & Technologies: HTML5, CSS, Git,, Splunk, ServiceNow Relevant Certifications

Generative AI Engineering with LLMs – Coursera 2

Generative AI for Data Scientists – Coursera 2

Experience

AI Engineer Intern

Bilvantis Technologies

Irving,TX

Sept 2024 – Apr 2025

Developed and deployed enterprise-grade RAG assistants using GPT-4, LangChain, and Vertex AI, enabling high-accuracy Q&A and document intelligence.

Built multi-agent systems using LangGraph + Autogen, coordinating LLMs across retrieval, reasoning, and summarization flows.

Engineered hybrid semantic search pipelines using Hugging Face embeddings with Pinecone and Weaviate, tuned for IT and regulatory knowledge bases.

Created custom chunking strategies with LlamaIndex (RecursiveTextSplitter) to preserve context in techni- cal documentation.

Designed and productionized LLM APIs using FastAPI and Node.js, integrated with internal IVR and enterprise platforms.

Developed an AI-powered migration tool that transforms Java Struts applications into FastAPI and React using OpenAI’s real-time API and Hugging Face models.

Led prompt engineering for zero-shot, few-shot, and CoT use cases to optimize customer support and compliance outputs.

Engineered a Streamlit-based document QA system using Mistral-7B embeddings with FAISS for semantic search and Hugging Face’s instruct models for answer generation

Prototyped agent orchestration workflows using LangChain, scaled them to production with LangGraph.

Developed a Twilio-based AI voice assistant using FastAPI, OpenAI’s real-time API, and Vosk for live 1

transcriptions.

Created internal demos, PoCs, and documentation for reusable GenAI components and LangChain-based templates.

Collaborated with cross-functional product teams to integrate LLM-driven insights into enterprise CRM and workflow systems.

AI Engineer

Infosys

Hyderabad, INDIA

Sep 2021 – July 2023

Developed LLM-based assistants using GPT-3 (Azure OpenAI) and LangChain, integrated into Microsoft Teams for HR and IT support automation.

Improved document search accuracy by 25% by deploying RAG pipelines using FAISS and Azure Cognitive Search with transformer embeddings.

Fine-tuned DistilBERT and GPT-2 for classification, intent detection, and document parsing tasks.

Created FastAPI microservices to serve LLM outputs and integrated them into enterprise dashboards and CRMs.

Designed and evaluated prompt strategies (zero-shot, few-shot, CoT) for summarization and internal support queries.

Prototyped LangGraph-style planners for multi-step task execution and agent coordination.

Containerized applications with Docker, deployed via Kubernetes, and monitored using CloudWatch and Grafana.

Developed internal prompt evaluation dashboards to assess LLM accuracy, latency, and cost across use cases.

Integrated LLM-enhanced workflows into legacy systems using REST APIs and message brokers (Kafka/Azure Queues).

Machine Learning Engineer

Avon Technologies Pvt Ltd

Hyderabad, INDIA

Mar 2020 – Aug 2021

Built NER and text classification models using BERT, spaCy, and TF-IDF for HR document parsing and support ticket routing.

Developed chatbot prototypes using Rasa and Dialogflow, enhanced with GPT-2 fallback generation for improved response coverage.

Created summarization tools using GPT-2, deployed as microservices to support internal search and knowl- edge workflows.

Exposed NLP results via Flask APIs, integrated with internal HR portals and analytics dashboards.

Containerized and deployed applications with Docker, automated releases via Jenkins for consistency and version control.

Integrated structured and unstructured training data using Azure SQL and Blob Storage, enabling scalable LLM fine-tuning.

Increased retrieval precision by designing keyword-to-prompt pipelines, boosting LLM query relevance by 20% in internal tools.

Collaborated with business analysts to map LLM outputs into structured formats for reporting and automa- tion triggers.

Education

Southeast Missouri State University

MS in Applied Computer Science

Aug 2023 – May 2025

St Martin’s Engineering College, Hyderabad, India

B.Tech in Computer Science and Engineering

June 2017 – May 2021

2

Contact this candidate