I’m an AI/ML Engineer with 5+ years of experience building production-ready machine learning and Generative AI systems, including LLM applications, RAG pipelines, recommendation models, and fraud detection. I enjoy taking models from experimentation to reliable deployment—focusing on measurable improvements in accuracy, retrieval quality, and end-to-end workflow efficiency. In my recent work, I’ve designed enterprise solutions that leverage Azure OpenAI and retrieval-augmented generation with hybrid retrieval and reranking, and I’ve implemented evaluation frameworks to assess groundedness, faithfulness, relevance, hallucinations, and latency. I also optimize inference performance through batching, caching, quantization, and throughput tuning, while building scalable pipelines and APIs with strong MLOps practices for consistent releases.…

Pranav A

I’m an AI/ML Engineer with 5+ years of experience building production-ready machine learning and Generative AI systems, including LLM applications, RAG pipelines, recommendation models, and fraud detection. I enjoy taking models from experimentation to reliable deployment—focusing on measurable improvements in accuracy, retrieval quality, and end-to-end workflow efficiency. In my recent work, I’ve designed enterprise solutions that leverage Azure OpenAI and retrieval-augmented generation with hybrid retrieval and reranking, and I’ve implemented evaluation frameworks to assess groundedness, faithfulness, relevance, hallucinations, and latency. I also optimize inference performance through batching, caching, quantization, and throughput tuning, while building scalable pipelines and APIs with strong MLOps practices for consistent releases.…

Available to hire

I’m an AI/ML Engineer with 5+ years of experience building production-ready machine learning and Generative AI systems, including LLM applications, RAG pipelines, recommendation models, and fraud detection. I enjoy taking models from experimentation to reliable deployment—focusing on measurable improvements in accuracy, retrieval quality, and end-to-end workflow efficiency.

In my recent work, I’ve designed enterprise solutions that leverage Azure OpenAI and retrieval-augmented generation with hybrid retrieval and reranking, and I’ve implemented evaluation frameworks to assess groundedness, faithfulness, relevance, hallucinations, and latency. I also optimize inference performance through batching, caching, quantization, and throughput tuning, while building scalable pipelines and APIs with strong MLOps practices for consistent releases.

See more

Skills

Experience Level

Language

Work Experience

AI/ML Engineer at Microsoft
October 1, 2024 - Present
Designed and developed enterprise AI/ML solutions using Python, PyTorch, scikit-learn, SQL, and Azure, improving model performance and workflow efficiency by 25%. Built Azure OpenAI and RAG-based applications with document ingestion, chunking, embeddings, hybrid retrieval, reranking, and context optimization using LangChain, LlamaIndex, FAISS, and vector search, improving retrieval relevance by 30%. Developed LLM evaluation pipelines using RAGAS, LangSmith, MLflow, and LLM-as-a-Judge to measure groundedness, faithfulness, relevance, hallucination, and latency, reducing manual evaluation effort by 40%. Implemented transformer-based NLP/LLM workflows with prompt engineering, structured outputs, and tool/function calling to improve response quality by 25%. Built scalable data pipelines (Spark, Airflow) and production AI services (FastAPI, Docker, Kubernetes) with 99%+ availability. Implemented MLOps and lifecycle workflows with MLflow and CI/CD, reducing deployment/release cycles by 35%,
Machine Learning Engineer at Amazon
January 1, 2020 - November 1, 2023
Designed and developed machine learning and statistical models using Python, SQL, pandas, NumPy, scikit-learn, XGBoost, and PySpark for large-scale datasets, improving predictive model performance by 18%. Built personalized product recommendation and ranking systems using collaborative filtering and content-based methods, increasing relevance and driving 12% improvement in click-through/conversion. Developed fraud detection and buyer-risk models using XGBoost, Random Forest, and anomaly detection, improving fraud detection precision by 16% while reducing false positives by 10%. Conducted advanced EDA, hypothesis testing, A/B testing, segmentation, and feature analysis, reducing manual analytical effort by 30%. Built scalable batch/distributed pipelines using PySpark/Spark/SQL and AWS (S3, EMR), reducing data processing time by 35%. Trained and evaluated models with cross-validation and hyperparameter optimization using ROC-AUC, precision, recall, and F1, improving stability/generalizat

Education

Master of Science in Data Science at University of Nebraska Omaha
January 11, 2030 - August 28, 2026

Qualifications

Add your qualifications or awards here.

Industry Experience

Software & Internet, Computers & Electronics, Professional Services

Skills

Experience Level