I build the entire stack on Databricks — streaming pipelines, production ML, and LLM-powered agents. From raw data to governed intelligence, in one platform.
Databricks · Data & AI
Most engineers own one slice of the data lifecycle. I own the whole medallion architecture — from ingestion through to the AI layer — so a project doesn't need five specialists to ship.
I'm a Data & AI Engineer who treats Databricks as home. Over five years I've shipped production statistical models, built enterprise-scale data platforms, and now engineer the full lakehouse — data, ML, and generative AI as one continuous system.
My edge is range with depth: I can stand up a streaming Delta pipeline, register and serve a model with MLflow, and wire an LLM agent over governed data — without handing off between four different roles.
I'm certified across the Databricks and Microsoft Fabric stacks, hold an MS in Artificial Intelligence from LUMS, and care about systems measured on real latency, real cost, and real users.
A conversational analytics agent that turns natural-language questions into governed SQL over Unity Catalog tables. Built on Databricks with a retrieval layer for schema grounding, a tool-calling agent for query planning, and Model Serving for deployment — so business users query the lakehouse in plain language.
A retrieval-augmented assistant over an enterprise document corpus, built on Databricks Vector Search and the Foundation Model APIs. Delivers grounded, cited answers with an MLflow-based evaluation harness measuring retrieval quality and answer faithfulness before anything ships.
An end-to-end MLOps pipeline on Databricks: features engineered into the Feature Store, training tracked in MLflow, models promoted through the Model Registry, and deployed to a serving endpoint with drift monitoring — retraining orchestrated by Databricks Workflows.
A production ingestion-to-gold pipeline using Delta Lake and Lakeflow declarative pipelines — streaming and batch through bronze, silver, and gold layers with data-quality expectations enforced at each hop.
Benchmarked six segmentation architectures for autonomous navigation and built a Bird's-Eye-View heading controller on the road mask — 82% mIoU at 88 FPS, with a real-time safety decision system.
Fine-tuned XLM-R and mBERT for English-to-Urdu transfer on a 21.7K-tweet corpus, with a fairness analysis exposing a 20× recall disparity across hate sub-categories.
A live delivery platform with a WhatsApp ordering bot powered by an LLM — a rule-based-first, AI-fallback design for reliable natural-language ordering, deployed and running for a real business.
He's a Data & AI Engineer specializing in the Databricks Lakehouse, building end-to-end systems across data engineering (Delta Lake pipelines), production machine learning (MLflow, Feature Store), and generative AI (RAG and LLM-powered agents) — backed by 5+ years across production data science and enterprise analytics.
Databricks is the primary platform — Delta Lake, Unity Catalog, MLflow, Mosaic AI, Vector Search, and Lakeflow pipelines. He also brings deep Microsoft data-platform experience with Fabric and Power BI, plus Azure, and works day to day in Python, PySpark, and SQL.
Yes — open to senior Data & AI and ML Engineering roles, including remote and Europe-based opportunities in energy, manufacturing, and industrial data. The fastest way to reach him is by email or LinkedIn, both linked below.
Open to senior Data & AI / ML Engineering roles — remote and across Europe. If you're building on the lakehouse, let's talk.