Data Scientist Ai Automation

Raghu Ram Ammula

***********@*****.*** 445-***-**** https://www.linkedin.com/in/raghu-ram-sai-ammula-a20860326/ SUMMARY

AI automation, retrieval, vector orchestrating goals Engineer to search reduce and with and multimodal agent (manual FAISS, 3+ document workflows, years effort, Chroma)agents of understanding. experience increase automating, using LangChain, Python, retention, building ML Designed Hugging FastAPI, lifecycles and production-drive and and Face, (CI/cloud-measurable deployed CD, and ready native MLflow, model LLM scalable impact. platforms and fine-Airflow)ML tuning pipelines (systems AWS,, and (LoRA, aligning Azure)for across real-QLoRA). Skilled AI customer time capabilities . in inference, Strong prompt intelligence, experience with engineering, RAG-business based voice in SKILLS

Programming Languages & Libraries, Tools: Python, SQL, Git, GitHub, REST APIs, Web Services, Scikit-learn, TensorFlow, PyTorch, Hugging Face, Transformers, LangChain, LangGraph, DVC, FastAPI, Flask, Tableau, Microsoft Excel NLP, LLMs & Generative AI: GPT-4, GPT-3.5, Claude, Mistral, Sentiment Analysis, Named Entity Recognition (NER), Summarization, Question Answering, OCR, RAG, Prompt Engineering, Fine-tuning (LoRA, QLoRA), Predictive Modeling, Feature Engineering, Hyperparameter Tuning, Problem Solving, Attention to Detail, Stakeholder Collaboration Cloud & MLOps: AWS (SageMaker, Lambda, Bedrock), GCP, Azure, Docker, Kubernetes, CI/CD, MLflow, Airflow, Prefect AI Workflow, Embeddings & Multimodal Tools: Multi-Agent Workflows, AI Orchestration, MCP, A2A, Workflow Automation, Streamlit, Hugging Face Spaces, FAISS, OpenAI Embeddings, ChromaDB, Whisper ASR, gTTS, TTS APIs, EXPERIENCE

VAZHRAA NIRMANN.PVT.LTD March 2021 – May 2023

Data Scientist Hyderabad, India

• Developed and deployed an XGBoost churn prediction model integrated into the CRM, enabling targeted retention campaigns that reduced churn by 8% across a customer base of over 10,000

• Spearheaded development of a real-time Power BI dashboard for marketing campaign analytics, reducing decision-making time by 40% and enhancing visibility across 5 business units.

• Built advanced SQL analytics tools to monitor supplier pricing patterns, driving data-informed negotiations and cutting procurement costs by 18%, yielding 60L+ annual savings.

• Designed and implemented a complete ML pipeline—from ingestion to hyperparameter tuning—boosting CRM predictive accuracy by 25%, leading to improved customer engagement strategies.

• Containerized ML models using Docker and orchestrated scalable production deployment on Kubernetes clusters, achieving 30% reduction in inference latency and 99.9% system uptime. Adroit March 2020 - March 2021

Data Analyst Hyderabad, India

• Increased forecast accuracy by 35% through the development of predictive models using Python and Scikit-learn, improving demand planning and resource allocation.

• Automated reporting workflows using Excel and Tableau across 4 cross-functional teams, resulting in 40% faster turnaround for performance insights and data access.

• Engineered and maintained robust ETL pipelines in Python and SQL, integrating data from 6+ disparate sources into a centralized data warehouse, increasing data availability and boosting query performance by 45%. PROJECTS

Resume-to-AI Chatbot (Agentic RAG Pipeline) May 2025

• Built a context-aware chatbot using OpenAI’s text-embedding-ada-002 + FAISS, enabling sub-100 ms semantic search on resume fragments.

• Designed self-reflective, retrieval, and critic agents for layered reasoning, significantly reducing hallucinations and improving Q&A accuracy.

• Developed and deployed a FastAPI microservice with Streamlit UI, fully containerized using Docker and hosted on Amazon Bedrock.

End-to-End AI Recruiting Agent with LLMs, Voice Interface, and Evaluation Automation June 2025

• Automated candidate sourcing using Tavily, RapidAPI, and GPT-4, boosting recruiter efficiency by 65%, reducing manual effort.

• Built a multimodal recruiter assistant with Mistral, Fast Whisper, and gTTS for voice-based AI interviews and real-time interactions.

• Orchestrated LangChain-style agent flows for personalized Q&A, automated scoring, and PDF feedback report generation, reducing manual evaluation effort by 70% and streamlining recruiter decision-making. Deployed in Azure AI Studio. Domain-Specific Llama Fine-Tuning for Customer Support June 2025

• • Fine-hallucination Deployed tuned real-LLaMA rate time by inference 2 37%with . 5K+ via FastAPI instruction-on Hugging response Face pairs Spaces via LoRA, using Docker, improving enabling response low-latency accuracy product by 48% query and resolution. lowering EDUCATION

Drexel University, Philadelphia, PA April 2025

Master of Science in Business Analytics (Statistics& Artificial Intelligence) (GPA 3.8)

Contact this candidate