Yaseer Ahmed Syed

Developed AI-powered incident analysis and automation workflows using Python, LLMs, RAG, and LangGraph to analyze production logs, identify critical error patterns, and retrieve historical incident resolutions. Built RAG pipelines with embeddings and contextual retrieval for incident insights and resolution recommendations. Implemented agentic workflows (LangChain/LangGraph) to orchestrate multi-step investigations, integrating MCP/tool calling with ServiceNow REST APIs for enriched operational context. Created log ingestion and preprocessing pipelines (Filebeat/Winlogbeat), and delivered scalable Python/Flask REST APIs for semantic retrieval and recommendations. Used Dynatrace and Azure Monitor for observability and root-cause analysis, improving operational reliability. Supported Docker deployments and CI/CD/LLMOps practices for production readiness and consistent releases.

Built and evaluated LSTM-based deep learning models for speculative purchasing pattern detection and time-series demand forecasting using historical healthcare purchasing data. Designed and maintained automated Dataiku ML pipelines for data ingestion, cleansing, transformation, validation, and preparation. Performed time-series feature engineering and compared LSTM vs Prophet forecasting models for trend analysis and predictive performance. Developed Python and Flask REST APIs to serve real-time model predictions to enterprise applications. Applied MLOps practices across training, validation, deployment, and monitoring. Optimized SQL queries/indexing and improved backend processing, achieving performance gains, and deployed ML-integrated services on cloud infrastructure in collaboration with cross-functional teams.

Developed Python/Flask prediction services by integrating serialized ML models into business applications using reusable API endpoints. Prepared structured SQL datasets, added input validation, transformed inference outputs into actionable analytics, and performed API/system integration testing. Supported backend automation, debugging, and documentation for ML-enabled reporting and operational workflows.