AI/ ML Engineer with hands-on industry experience designing and deploying end-to-end machine learning and LLM-powered systems and intelligent AI agents. Skilled in building scalable data pipelines, feature engineering wo…
AI/ ML Engineer with hands-on industry experience designing and deploying end-to-end machine learning and LLM-powered systems and intelligent AI agents. Skilled in building scalable data pipelines, feature engineering workflows, predictive models, and enterprise-grade Retrieval-Augmented Generation (RAG) architectures with hybrid search and re-ranking. Strong expertise in Python, deep learning frameworks, backend API development, and database integration. Experienced with Docker-based deployment, cloud ML environments, and MLOps practices including experiment tracking, logging, monitoring, and model lifecycle management.
Designed and implemented a multimodal RAG system processing 1,000+ document chunks to answer domain-specific queries from complex PDF documents.
Built Retrieval-Augmented Generation (RAG) pipelines using LangChain, ingestion, document chunking, embedding generation, vector indexing, hybrid search, and re-ranking to improve retrieval accuracy.
Integrated structured databases (PostgreSQL/MySQL) for document metadata management and dynamic query handling.
Engineered advanced RAG reasoning logic to minimize hallucinations and improve contextual answer consistency.
Refined LLM-based SQL query generation using prompt engineering techniques for accurate database retrieval.
Expanded RESTful APIs using FastAPI, enabling real-time query responses for AI-driven applications.
Enhanced structured logging, monitoring, and error-handling mechanisms to improve system reliability and debugging.
Developed machine learning forecasting models to predict product demand using historical sales time-series data.
Conducted exploratory data analysis and identified highly intermittent demand patterns with a large proportion of zero-sales periods.
Performed time-series anomaly detection to identify irregular demand spikes and outliers.
Generated predictive features including lag variables, rolling statistics, and temporal indicators to capture demand trends.
Applied forecasting algorithms including XGBoost, Random Forest, and Croston methods to model intermittent demand behavior.
Constructed a stacking ensemble model combining multiple algorithms to improve prediction stability and robustness.
Assessed forecasting performance using MAE, RMSE, and R2 evaluation metrics.
Implemented recursive forecasting techniques to generate multi-step future demand predictions.
Designed and implemented a multimodal RAG system processing 1,000+ document chunks to answer domain-specific queries from complex PDF documents., Built Retrieval-Augmented Generation (RAG) pipelines using LangChain, ingestion, document chunking, embedding generation, vector indexing, hybrid search, and re-ranking to improve retrieval accuracy., Integrated structured databases (PostgreSQL/MySQL) for document metadata management and dynamic query handling., Engineered advanced RAG reasoning logic to minimize hallucinations and improve contextual answer consistency., Refined LLM-based SQL query generation using prompt engineering techniques for accurate database retrieval., Expanded RESTful APIs using FastAPI, enabling real-time query responses for AI-driven applications., Enhanced structured logging, monitoring, and error-handling mechanisms to improve system reliability and debugging.
• Architected a multimodal Retrieval-Augmented Generation (RAG) system capable of answering domain-specific queries from complex PDF documents containing text and tabular data, processing 1,000+ document chunks.• Constructed end-to-end RAG pipelines with LangChain, including document ingestion, chunking, embedding generation, vector indexing, hybrid search, and re-ranking to improve retrieval relevance.• Integrated PostgreSQL and MySQL databases to manage document metadata and support scalable query processing.• Implemented advanced reasoning logic within the RAG pipeline to reduce hallucinations and improve contextual answer consistency.• Optimized LLM-driven SQL query generation through prompt engineering techniques for accurate database retrieval.• Developed FastAPI-based REST services enabling real-time responses for AI-powered applications.• Established structured logging, monitoring, and error-handling pipelines to enhance system reliability and simplify debugging.
• Designed machine learning models to forecast product demand from historical sales time-series data.• Conducted exploratory data analysis identifying highly intermittent demand patterns with a significant proportion of zero-sales periods.• Performed time-series anomaly detection to identify irregular demand spikes and anomalies.• Generated predictive features including lag variables, rolling statistics, and seasonal indicators to capture temporal patterns.• Applied forecasting algorithms including XGBoost, Random Forest, and Croston methods to model intermittent demand.• Constructed a stacking ensemble framework combining multiple algorithms to improve prediction stability and generalization.• Assessed model performance using MAE, RMSE, and R² metrics to validate predictive accuracy.• Implemented recursive forecasting strategies to generate multi-step future demand predictions.
The AI-first Recruitment Operating System that unifies and automates the entire hiring lifecycle for modern enterprises worldwide.
Dr Job FZ LLC. 2026 © All Rights Reserved
