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

I am an AI/ML Engineer with 3+ years of experience designing and deploying scalable ML, NLP, and Generative AI solutions across mortgage finance, semiconductor manufacturing, and enterprise analytics. I build cloud-native ML platforms with Python, Azure ML, AWS, and MLOps best practices to deliver production-grade models, explainability, and robust monitoring in regulated environments. I thrive at turning data into measurable business impact, from LLM-powered document intelligence for mortgage underwriting to predictive ML solutions for manufacturing. I enjoy end-to-end model lifecycle management, prompt engineering for reliable AI workflows, and collaborating with cross-functional teams to deliver auditable, scalable data products.…I am an AI/ML Engineer with 3+ years of experience designing and deploying scalable ML, NLP, and Generative AI solutions across mortgage finance, semiconductor manufacturing, and enterprise analytics. I build cloud-native ML platforms with Python, Azure ML, AWS, and MLOps best practices to deliver production-grade models, explainability, and robust monitoring in regulated environments. I thrive at turning data into measurable business impact, from LLM-powered document intelligence for mortgage underwriting to predictive ML solutions for manufacturing. I enjoy end-to-end model lifecycle management, prompt engineering for reliable AI workflows, and collaborating with cross-functional teams to deliver auditable, scalable data products.

Hrithik Puri





I am an AI/ML Engineer with 3+ years of experience designing and deploying scalable ML, NLP, and Generative AI solutions across mortgage finance, semiconductor manufacturing, and enterprise analytics. I build cloud-native ML platforms with Python, Azure ML, AWS, and MLOps best practices to deliver production-grade models, explainability, and robust monitoring in regulated environments. I thrive at turning data into measurable business impact, from LLM-powered document intelligence for mortgage underwriting to predictive ML solutions for manufacturing. I enjoy end-to-end model lifecycle management, prompt engineering for reliable AI workflows, and collaborating with cross-functional teams to deliver auditable, scalable data products.…I am an AI/ML Engineer with 3+ years of experience designing and deploying scalable ML, NLP, and Generative AI solutions across mortgage finance, semiconductor manufacturing, and enterprise analytics. I build cloud-native ML platforms with Python, Azure ML, AWS, and MLOps best practices to deliver production-grade models, explainability, and robust monitoring in regulated environments. I thrive at turning data into measurable business impact, from LLM-powered document intelligence for mortgage underwriting to predictive ML solutions for manufacturing. I enjoy end-to-end model lifecycle management, prompt engineering for reliable AI workflows, and collaborating with cross-functional teams to deliver auditable, scalable data products.

Available to hire

I am an AI/ML Engineer with 3+ years of experience designing and deploying scalable ML, NLP, and Generative AI solutions across mortgage finance, semiconductor manufacturing, and enterprise analytics. I build cloud-native ML platforms with Python, Azure ML, AWS, and MLOps best practices to deliver production-grade models, explainability, and robust monitoring in regulated environments.

I thrive at turning data into measurable business impact, from LLM-powered document intelligence for mortgage underwriting to predictive ML solutions for manufacturing. I enjoy end-to-end model lifecycle management, prompt engineering for reliable AI workflows, and collaborating with cross-functional teams to deliver auditable, scalable data products.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

English

Fluent

Work Experience

AI/ML Engineer at Freddie Mac

November 1, 2024 - Present

Designed and implemented an LLM-powered document intelligence pipeline using Python, NLP preprocessing, LangChain, embeddings, and vector databases, automating mortgage document understanding and reducing manual underwriting review effort by 40%. Built a retrieval-augmented generation (RAG) framework integrating verified mortgage data sources with LLMs, improving response accuracy, auditability, and regulatory safety for AI-assisted underwriting workflows. Optimized prompt engineering and response validation logic to generate explainable risk summaries, enabling under writers to identify income, disclosure, and compliance risk signals with 35% faster decision cycles. Developed supervised credit risk models using XGBoost, LightGBM, and Scikit-learn on borrower, property, and macroeconomic features, achieving strong discriminatory power for probability-of-default prediction. Engineered cloud-native ML pipelines on Microsoft Azure using Azure Data Factory, Synapse Analytics, and Azure Fea

Data Scientist (Machine Learning) at Veeco Instruments

July 1, 2023 - December 1, 2023

Developed predictive machine learning models using XGBoost, LightGBM, and Scikit-learn on high-frequency semiconductor equipment telemetry, improving early fault detection recall by 90%+ and reducing yield loss risk. Engineered advanced time-series and process-level features (rolling statistics, lag variables, anomaly indicators) with Python, NumPy, and SQL, increasing model precision by 18% over rule-based baselines. Applied explainable AI techniques using SHAP to interpret model outputs, enabling process engineers to identify root causes of yield degradation and accelerating corrective action cycles by 35%. Operationalized reproducible ML workflows using MLflow, Docker, and AWS EC2, supporting controlled experimentation, versioned model tracking, and scalable deployment for manufacturing analytics use cases.

Data Engineer at Persistent Systems

January 1, 2021 - July 1, 2022

Architected scalable batch and streaming data pipelines using Python, PySpark, Apache Spark, and Airflow, processing millions of finance and e-commerce records daily and reducing data availability latency by 48%. Unified transactional, clickstream, and customer datasets within a cloud data lake architecture (AWS S3, Glue, Athena, Redshift), enabling cross-domain analytics and improving enterprise reporting consistency by 35%. Engineered analytics-optimized fact–dimension models and partitioned datasets to support BI and ML workloads, accelerating complex SQL queries and dashboard performance by 45%. Implemented real-time ingestion frameworks using Apache Kafka and AWS Kinesis to capture high-velocity events, supporting near real-time risk monitoring and customer behavior analysis. Delivered ML-ready feature datasets for fraud detection, credit risk, and demand forecasting models, reducing model data preparation cycles by 30% and improving experimentation velocity. Enabled self-servic