Skip to content
Uzair Khan
Data & AI EngineerRemote

I build the data platforms that power AI.

4+ years in production. I design the data pipelines and lakehouses on Azure, Snowflake, Databricks and AWS, then build the RAG, ML and LLM systems that run on them.

Book a callView work
Open to senior Data & AI roles + select consulting
Data Engineer @ WeCrunch (UAE)· ex-OctDaily · ex-Create Impact
  • 4+ yrs in production
  • ·Azure · Snowflake · Databricks · AWS
  • ·Healthcare & FMCG
Certified byIBM · DeepLearning.AI · Snowflake · Databricks
How I work

From raw data to shipped intelligence

Great AI needs great data engineering. I own the whole path, from the pipelines that feed the model to the AI layer your users actually touch.

  1. SourcesDBs · APIs · Files
  2. IngestKafka · ADF · Logic Apps
  3. LakehouseDelta · Snowflake
  4. TransformPySpark · dbt · DQ
  5. AI LayerRAG · Agents · ML
  6. ProductApps · Dashboards
Expertise

What I do

Two disciplines, one through-line: the data foundation and the intelligence built on top of it.

Data Engineering

Production lakehouses & pipelines that are reliable, auditable, and cheap to run.

  • Azure: ADF, Databricks, Delta Lake, Event Grid, Logic Apps
  • Snowflake: Snowpipe, Streams & Tasks, SnowSQL, Snowpark
  • AWS: Glue, Lambda, S3, Lake Formation, Step Functions
  • Medallion architecture · ELT orchestration · PySpark · Airflow

AI Engineering

LLM and ML systems built on solid data foundations, from RAG to clinical decision support.

  • RAG pipelines & LLM apps (Llama, Pinecone, vector search)
  • NLP and ML: recommendations, forecasting, classification
  • Healthcare AI: clinical decision support, EHR/EMR modeling
  • From prototype to production, with the data plumbing to match

Data Quality & Governance

Trust by design.

  • Great Expectations DQ frameworks
  • Schema-drift detection & alerting
  • Lineage, audit logging, access control

Platform & DevOps

Ship it like software.

  • CI/CD with GitHub Actions
  • Config-driven, idempotent pipelines
  • Cost & cluster optimization

Teaching & Mentoring

I mentor peers and students on cloud data engineering. Explaining things clearly is a senior skill.

Selected work

Systems that shipped, and the numbers to prove it

Production builds across FMCG, healthcare, and analytics. Real platforms running in real companies, not demos.

WeCrunch
Enterprise

FMCG Multi-Country Data Platform

Autonomous SharePoint ingestion (Logic Apps + Blob + Delta) and idempotent Bronze/Silver/Gold pipelines feeding near-real-time market-share reporting for a multi-country FMCG strategy.

70%faster pipeline triggers
AzureDatabricksDelta LakeEvent GridPySparkPower BI
OctDaily
AI

Clinical AI & EHR Data Platform

Refactored fragmented EHR/clinical databases into normalized schemas and built ELT on ADF + Databricks, powering a clinical decision-support module for medicine recommendations and patient-history visualization.

+30%data-flow throughput
AzureADFDatabricksADLSSQLML
Data Engineering

Real-Time Sales Analytics Pipeline

Event-driven AWS pipeline (Glue, Lambda, S3, EventBridge) with a Raw → Staging → Curated lake via Lake Formation, and a metadata-driven Step Functions workflow handling retries, alerts, and dependencies.

-45%processing latency
AWS GlueLambdaS3EventBridgeStep FunctionsSNS
Data Engineering

Snowflake ELT with CI/CD & Data Quality

A production Snowflake ELT pipeline built with SnowSQL and the Snowpark Python API, fronted by a Great Expectations data-quality gate and GitHub Actions CI/CD for version-controlled migrations. Lineage and audit logging via Streams kept it audit-ready from day one.

0downtime schema changes
SnowflakeSnowparkSnowSQLGreat ExpectationsGitHub Actions
View all work
Years building production data systems
Faster pipeline triggers (WeCrunch ingestion framework)
Lower monthly cloud compute & storage cost
of 300 teams at the UBL Datathon 2022
Architecture

I think in systems, not scripts

The architectures I design and ship, from RAG pipelines and multi-agent systems to medallion lakehouses and cloud data platforms. Explore each one as an interactive, expandable diagram.

RAG PipelineMulti-Agent SystemsMedallion LakehouseSnowflake WarehouseAzure StreamingAWS Event-Driven
RAG pipeline live pattern
  1. User querynatural language
  2. Retrieve contextVector DB · Lakehouse
  3. LLM reasoningLlama · grounded
  4. Grounded answercited · accurate
Trust

Backed by results and references

What collaborators say, alongside the credentials behind the work.

“Uzair rebuilt our ingestion layer end to end. Pipelines that used to need babysitting now just run, and our dashboards refresh in minutes instead of hours.”
EEngineering Lead, FMCG analytics
“Rare combination: he can architect the data platform and ship the AI on top of it. He bridged our engineering and clinical teams without friction.”
PProduct Lead, Healthcare AI
“He cut our Databricks bill by a third without touching a single dashboard's SLA. He knows exactly where the money leaks in a pipeline.”
HHead of Data, Retail analytics
“Onboarded, understood our Snowflake mess, and shipped a clean medallion model in weeks. Communicates with non-technical stakeholders better than most leads.”
CCTO, Early-stage startup
“I learned more about real-world data engineering from his mentorship than from any course. He is patient, precise, and genuinely invested in how you grow.”
DData Engineer, former mentee

Certifications

  • IBM Data Science
  • IBM Data Engineering
  • DeepLearning.AI NLP Specialization
  • Snowflake Hands-On Essentials (×3)
  • Databricks Lakeflow Spark Pipelines
  • 365DataScience SQL & Advanced SQL

Awards

  • Runner-Up, UBL Datathon · 20222nd of 300 teams · AI Instant-Decision Loan Portal
  • Runner-Up, IEEE NEDUET DS Competition · 20212nd of 30 teams · ATM-downtime ML model
About

I'm a Data & AI Engineer who builds both the data platforms and the AI that runs on them.

I came up through data science and national competitions, then spent the last four years deep in production data engineering. I've built pipelines, lakehouses and warehouses on Azure, Snowflake, Databricks and AWS for healthcare and FMCG teams, plus the RAG, ML and decision-support systems that sit on top. The pattern I keep seeing is simple: AI only works when the data underneath it does, so I build both layers.

  1. 2021Data Engineer· Create Impact

    AWS Glue/Lambda ETL · Snowflake models

  2. 2024AI Engineer· OctDaily (USA)

    Healthcare AI · EHR/EMR ELT · decision support

  3. 2024 →Data Engineer· WeCrunch (UAE)

    FMCG multi-country lakehouse · streaming

Muhammad Uzair Khan, Data and AI Engineer
Uzair KhanData & AI

Let's build something that ships.

Hiring for a senior Data/AI role, or need a data platform that actually holds up in production? Let's talk.

Book a callDownload résumé
or email me directly at muhammaduzairkhan329@gmail.com