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

AI/ML Engineer with 7+ years of hands-on experience building and shipping Generative AI and machine learning systems across legal tech, fintech, healthcare, and industrial domains. I turn complex, messy business problems into working AI—e.g., RAG pipelines querying millions of documents, multi-agent systems that draft contracts autonomously, and real-time NLP pipelines for adverse drug event detection. I work across the full stack: data wrangling, model training, fine-tuning, API development, cloud deployment, and production monitoring with logs, metrics, and dashboards. I’ve fine-tuned LLMs using LoRA/QLoRA/PEFT, built agentic systems with LangChain/LangGraph/CrewAI, and implemented LLMOps/observability and governance controls to keep AI behavior reliable in production.…AI/ML Engineer with 7+ years of hands-on experience building and shipping Generative AI and machine learning systems across legal tech, fintech, healthcare, and industrial domains. I turn complex, messy business problems into working AI—e.g., RAG pipelines querying millions of documents, multi-agent systems that draft contracts autonomously, and real-time NLP pipelines for adverse drug event detection. I work across the full stack: data wrangling, model training, fine-tuning, API development, cloud deployment, and production monitoring with logs, metrics, and dashboards. I’ve fine-tuned LLMs using LoRA/QLoRA/PEFT, built agentic systems with LangChain/LangGraph/CrewAI, and implemented LLMOps/observability and governance controls to keep AI behavior reliable in production.

Apuroop Yarabarla

AI Engineer, Data Scientist, Software Engineer, +1





AI/ML Engineer with 7+ years of hands-on experience building and shipping Generative AI and machine learning systems across legal tech, fintech, healthcare, and industrial domains. I turn complex, messy business problems into working AI—e.g., RAG pipelines querying millions of documents, multi-agent systems that draft contracts autonomously, and real-time NLP pipelines for adverse drug event detection. I work across the full stack: data wrangling, model training, fine-tuning, API development, cloud deployment, and production monitoring with logs, metrics, and dashboards. I’ve fine-tuned LLMs using LoRA/QLoRA/PEFT, built agentic systems with LangChain/LangGraph/CrewAI, and implemented LLMOps/observability and governance controls to keep AI behavior reliable in production.…AI/ML Engineer with 7+ years of hands-on experience building and shipping Generative AI and machine learning systems across legal tech, fintech, healthcare, and industrial domains. I turn complex, messy business problems into working AI—e.g., RAG pipelines querying millions of documents, multi-agent systems that draft contracts autonomously, and real-time NLP pipelines for adverse drug event detection. I work across the full stack: data wrangling, model training, fine-tuning, API development, cloud deployment, and production monitoring with logs, metrics, and dashboards. I’ve fine-tuned LLMs using LoRA/QLoRA/PEFT, built agentic systems with LangChain/LangGraph/CrewAI, and implemented LLMOps/observability and governance controls to keep AI behavior reliable in production.

Available to hire

AI/ML Engineer with 7+ years of hands-on experience building and shipping Generative AI and machine learning systems across legal tech, fintech, healthcare, and industrial domains. I turn complex, messy business problems into working AI—e.g., RAG pipelines querying millions of documents, multi-agent systems that draft contracts autonomously, and real-time NLP pipelines for adverse drug event detection.

I work across the full stack: data wrangling, model training, fine-tuning, API development, cloud deployment, and production monitoring with logs, metrics, and dashboards. I’ve fine-tuned LLMs using LoRA/QLoRA/PEFT, built agentic systems with LangChain/LangGraph/CrewAI, and implemented LLMOps/observability and governance controls to keep AI behavior reliable in production.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

Work Experience

AI Engineer at UPS

January 1, 2025 - Present

Architected and deployed a production-grade LLM-powered GenAI automation platform using RAG, agentic workflows, and tool calling to automate high-volume business query resolution and operational decision support. Built a RAG-based knowledge assistant over enterprise FAQs, policies, process documents, and internal knowledge bases to generate grounded citation-backed responses. Engineered a Dockerized multi-agent system using LangGraph/StateGraph to orchestrate intent classification, retrieval, policy validation, tool execution, response generation, and human escalation. Implemented parallel agent execution flows to reduce orchestration latency. Designed MCP-style plug-and-play tool interfaces for knowledge search, record lookup, policy validation, CRM updates, and ticket creation. Integrated LLMs with context injection, advanced prompt engineering, structured outputs (Pydantic JSON schemas), and deterministic routing. Implemented confidence-based escalation, hallucination checks, docume

Data Scientist at Pragma Info Systems

September 1, 2023 - December 1, 2024

Built Intelligent Document Comparison systems to analyze, align, and surface meaningful differences across policy documents, contracts, regulatory filings, and product specifications using LLM-powered strategies. Implemented LangChain/LangGraph workflows with stateful orchestration for parsing, entity extraction, graph linking, section alignment, clause-level diffing, and difference summarization. Used Neo4j as a knowledge graph layer to model relationships between clauses, defined terms, cross-references, and entities so differences were meaningful to human reviewers. Layered RAG for contextual lookups when clauses/terms were ambiguous, including support knowledge base lookups, classification-based routing, and sentiment analysis for tickets. Maintained FastAPI endpoints for low-latency access. Performed prompt engineering for defensible, consistent legal-difference outputs, including material vs non-material change handling. Set up AWS SageMaker/Databricks pipelines for embedding mod

Software Developer at Intellectuals AI

July 1, 2021 - February 1, 2022

Built an end-to-end Voice-to-Text pipeline for ingesting call recordings and transcribing audio at scale, then applied NLP/knowledge-graph-based semantic processing to extract structure from transcripts. Implemented entity extraction and sentiment/intent/urgency/outcome classification for transcript tagging, including clustering/segmentation and churn prediction. Addressed imbalanced datasets with resampling, and performed data cleaning using SQL plus NLP/text mining. Developed forecasting models (time series call volume/demand forecasting, regression escalation likelihood) and extended predictive analytics to sales forecasting, dynamic/pricing optimization, and revenue optimization using XGBoost/LightGBM. Built ETL into MongoDB to move audio metadata, transcripts, and model outputs reliably across heterogeneous sources. Produced dashboards and reports in Tableau/Power BI and supported Agile workflows with mappings/test cases; also implemented deep learning (CNN/ANN, Keras/TensorFlow,

Software Developer at Accenture

January 1, 2018 - June 1, 2021

Developed Python/R production models for sales forecasting and price prediction, including classical ML (regression, random forests, gradient boosting, clustering) combined with deep learning where non-linear pricing interactions required it. Built an RPA layer for pulling structured/semi-structured inputs into prediction pipelines and structured text fields with NLP tools (e.g., NLTK). Owned end-to-end data flows: extraction, transformation/field mapping in Pandas, loading into targets, and generating outputs/reports. Produced recurring analytical reporting (time series trends, price modeling, sales forecasts, trend mapping) and built Tableau/Power BI dashboards to make forecasts usable for non-technical stakeholders. Implemented predictive analytics powering segmentation, price elasticity, demand forecasting, and pricing optimization; validated approaches through cohort-based A/B testing.