Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Full-stack AI/ML professional with 4+ years of experience designing, deploying, and operating production AI systems across LLMOps, NLP, machine learning, and data engineering. Proficient across the full AI/ML lifecycle—from research, development, and evaluation frameworks to scalable pipeline architecture and production delivery—owning end-to-end solutions independently. Hands-on expertise in LLMs (Claude 3, GPT, Azure OpenAI), RAG pipelines, evaluation (including BERTScore), hallucination mitigation, and orchestration using AWS Step Functions, SageMaker, and LangChain/LangGraph. Certified holder of AWS Machine Learning – Specialty and experienced across AWS, Azure, and GCP with strong MLOps, CI/CD, and automation capabilities.…Full-stack AI/ML professional with 4+ years of experience designing, deploying, and operating production AI systems across LLMOps, NLP, machine learning, and data engineering. Proficient across the full AI/ML lifecycle—from research, development, and evaluation frameworks to scalable pipeline architecture and production delivery—owning end-to-end solutions independently. Hands-on expertise in LLMs (Claude 3, GPT, Azure OpenAI), RAG pipelines, evaluation (including BERTScore), hallucination mitigation, and orchestration using AWS Step Functions, SageMaker, and LangChain/LangGraph. Certified holder of AWS Machine Learning – Specialty and experienced across AWS, Azure, and GCP with strong MLOps, CI/CD, and automation capabilities.

Vasantha Lakshmi Eda

Data Scientist, AI Engineer, Full Stack Developer, +2





Full-stack AI/ML professional with 4+ years of experience designing, deploying, and operating production AI systems across LLMOps, NLP, machine learning, and data engineering. Proficient across the full AI/ML lifecycle—from research, development, and evaluation frameworks to scalable pipeline architecture and production delivery—owning end-to-end solutions independently. Hands-on expertise in LLMs (Claude 3, GPT, Azure OpenAI), RAG pipelines, evaluation (including BERTScore), hallucination mitigation, and orchestration using AWS Step Functions, SageMaker, and LangChain/LangGraph. Certified holder of AWS Machine Learning – Specialty and experienced across AWS, Azure, and GCP with strong MLOps, CI/CD, and automation capabilities.…Full-stack AI/ML professional with 4+ years of experience designing, deploying, and operating production AI systems across LLMOps, NLP, machine learning, and data engineering. Proficient across the full AI/ML lifecycle—from research, development, and evaluation frameworks to scalable pipeline architecture and production delivery—owning end-to-end solutions independently. Hands-on expertise in LLMs (Claude 3, GPT, Azure OpenAI), RAG pipelines, evaluation (including BERTScore), hallucination mitigation, and orchestration using AWS Step Functions, SageMaker, and LangChain/LangGraph. Certified holder of AWS Machine Learning – Specialty and experienced across AWS, Azure, and GCP with strong MLOps, CI/CD, and automation capabilities.

Available to hire

Full-stack AI/ML professional with 4+ years of experience designing, deploying, and operating production AI systems across LLMOps, NLP, machine learning, and data engineering. Proficient across the full AI/ML lifecycle—from research, development, and evaluation frameworks to scalable pipeline architecture and production delivery—owning end-to-end solutions independently.

Hands-on expertise in LLMs (Claude 3, GPT, Azure OpenAI), RAG pipelines, evaluation (including BERTScore), hallucination mitigation, and orchestration using AWS Step Functions, SageMaker, and LangChain/LangGraph. Certified holder of AWS Machine Learning – Specialty and experienced across AWS, Azure, and GCP with strong MLOps, CI/CD, and automation capabilities.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Language

Work Experience

Data Scientist II at Amazon

October 1, 2025 - April 30, 2026

Architected an end-to-end LLM-powered insights automation system for monthly/weekly/daily issue lifecycle reporting, reducing manual reporting effort by ~80%. Designed and orchestrated scalable dual-package data pipelines (CDK-based infrastructure + Python science packages) and built an end-to-end inference pipeline using AWS Step Functions to validate, transform, generate top contributors, run Bedrock (Claude 3) inference, evaluate outputs, and write results to S3. Developed persona-based actionable recommendations, implemented evaluation frameworks (ground-truth alignment using BERT-based similarity, field completeness, and data quality checks) achieving 93% reliability, and deployed a Slack bot for automated delivery to 20+ MBR metrics. Also automated customer survey analysis pipelines for 7+ tools using Python, NLTK, and spaCy to produce structured CSAT insights and dashboards for leadership using Amazon QuickSight.

Agentic AI Prompt Engineer at CMG Financial

July 1, 2025 - October 1, 2025

Engineered a RAG-based mortgage policy search system using LangChain, Azure OpenAI, and Azure AI Search indexing 40+ policy documents to enable natural-language policy queries. Launched an AI paydown simulator as a chat interface using Python, RAG, and n8n to model loan repayment scenarios for 100+ end users. Independently delivered meeting transcript automation using event-driven architecture (Drive recordings listener, transcript scraping, OpenAI-based extraction/summarization). Built end-to-end automation workflows using Microsoft Power Automate and n8n and integrated AI applications into frontend interfaces.

Data Engineer at GoTackle (Powered by GoLogic)

January 1, 2025 - June 30, 2025

Designed and implemented serverless data pipelines using AWS Lambda, Redshift, and BigQuery for real-time high-volume business data. Built AI-powered business insight engines using OpenAI GPT, Vertex AI, and LangChain to generate automated recommendations and reports. Automated data transformations using dbt and Airflow and engineered event-driven workflows with EventBridge and DynamoDB to improve scalability and reliability. Collaborated to integrate AI-driven insights into user dashboards for 10,000+ SMB users.

Research Assistant at Arizona State University

January 1, 2023 - December 31, 2024

Built a RAG pipeline leveraging knowledge graph triplets for semantic search and contextual recommendations, improving retrieval accuracy by 35%. Fine-tuned pre-trained models (BERT, GPT-2) for domain-specific QA and summarization using Hugging Face Transformers. Developed NLP pipelines with spaCy/Transformers/custom tokenizers to extract structured data from 10M+ unstructured records, reducing manual review by 40%. Performed data cleaning on 500,000+ records improving data quality by 30% and delivered insights via 10+ visualizations. Developed an ML matching model improving accuracy by 25% and reducing reconciliation effort by 60%.

Data Science Analyst at Cardlytics

July 1, 2021 - November 30, 2022

Built a Streamlit application to aggregate and standardize merchant location data across sources, improving processing efficiency by 4x and reducing manual effort from 5 FTEs to 1. Developed a REST API for merchant location cleaning (MLC) to normalize and extract phone numbers from transaction data. Enhanced time-series forecasting using XGBoost, LSTM, and LightGBM, reducing operational risk by 25%. Deployed ML models via FastAPI and Docker. Optimized large-scale SQL queries for financial extraction by 25%, streamlined reporting with SAP BusinessObjects (30% faster), and implemented MLOps and CI/CD with Azure DevOps. Built batch and real-time ETL pipelines using Python, PySpark, AWS, and Kafka (15% speedup) and automated batch processing (50% workload reduction).

Data Science Analyst Intern (Volunteer) at Cardlytics

February 1, 2021 - June 30, 2021

Designed an unsupervised clustering model to group millions of transaction records using fuzzy matching and n-gram analysis to map transactions to likely merchant brands with OpenStreetMap data. Contributed to documentation, code reviews, and knowledge sharing. Built automated pipelines using Python, web scraping, and Google search APIs to create a location corpus mapping cities to zip codes. Preprocessed large datasets with NumPy/Pandas/Scikit-learn to support model quality.