Available to hire
I am Lakshmi Vyshnavi Ummadisetti, an AI/ML Engineer with over 4.5 years of experience delivering data-driven solutions across financial services and healthcare. I specialize in building scalable ML systems using Python, PyTorch, TensorFlow, and cloud-based MLOps to deliver measurable business impact.
My focus areas include feature engineering, model explainability, retrieval-augmented generation, and end-to-end AI/ML pipelines. I enjoy collaborating with cross-functional teams to translate complex problems into deployable solutions that improve risk management, compliance, and operational efficiency.
Skills
Language
English
Fluent
Work Experience
AI/ML Engineer at Citi
September 1, 2024 - PresentBuilt and deployed graph-based fraud detection models using NetworkX and transaction relationship networks, uncovering hidden risk links across accounts and reducing manual investigation load by 2,500+ cases per quarter. Developed retrieval-augmented generation (RAG) pipelines using Pinecone vector DB and transformer embeddings to answer complex regulatory and compliance queries, enabling analysts to retrieve context-rich information from over 3 million internal documents in under 5 seconds. Applied generative AI models (LLaMA, Hugging Face Transformers) to automate summarization of financial reports, internal communications, and client interactions, saving compliance teams 80+ hours per month in manual review. Engineered LLM-powered NLP solutions for analyzing unstructured customer support logs, emails, and chat transcripts, extracting risk indicators and compliance flags with real-time scoring. Implemented cloud-based ML/LLM pipelines using AWS SageMaker, Docker, and Kubernetes, auto
Machine Learning Engineer at Accenture
July 1, 2020 - August 1, 2023Constructed and deployed machine learning models for patient risk stratification and hospital readmission using scikit-learn, leveraging structured EHR, claims, lab, and demographic data, improving early-risk identification accuracy by 18%. Designed end-to-end data pipelines for multi-source healthcare datasets using Pandas, SQL, Apache Kafka, and cloud storage (AWS S3), reducing data processing time by 35% and improving analytical readiness. Architected supervised and unsupervised models including logistic regression, KNN, random forest, XGBoost, and clustering techniques to support population health analytics and utilization management across payer and provider use cases. Orchestrated advanced feature engineering on longitudinal patient data using NumPy, applying temporal aggregation, normalization, and clinical event encoding to improve model performance metrics (AUC / F1) by 12–15%. Formed and evaluated deep learning prototypes using TensorFlow and PyTorch, including LSTM/GRU seq
Education
Master of Science in Computer Science (Data Science track) at University of Texas at Dallas
January 11, 2030 - February 16, 2026Qualifications
Industry Experience
Financial Services, Healthcare
Skills
Hire a AI Engineer
We have the best ai engineer experts on Twine. Hire a ai engineer in Irving today.