The leader track 🤝 · for C-level, managers & data leads · 6 x 45 min · no code, ever
Thinking mode: what DataOps is, why fragile data costs money, how mature your org is, and how to make the build-buy-fund calls - each session ends with questions to ask your data team on Monday.
Why DataOps? From artisanal to industrialized data
The reliability tax fragile pipelines charge you, and the three lineages - DevOps, Agile, lean - that DataOps braids together.
🎯 Leader 2 · easyFour pillars, four business risks
DevOps, data CI/CD, MLOps and DBOps - each framed as the dollar risk it retires and the question it lets you ask.
📈 Leader 3 · mediumThe maturity model and the metrics that matter
Five levels from heroic-and-fragile to self-service, plus DORA-for-data: the four numbers that tell the truth.
💸 Leader 4 · decidingBuild, buy, or open-source: the platform bet
OSS stack vs Databricks/Snowflake/AWS, real cost ranges, and the case for one governed platform over a tool zoo.
🧩 Leader 5 · mediumWho owns what: org design and data contracts
Central team vs data mesh, team topologies, and data contracts as the social technology that breaks silos.
📊 Leader 6 · decidingProving it: ROI and a 90-day roadmap
The honest business case, the metrics a board will believe, and a phased plan that ships value in the first quarter.
The builder track 🛠️ · for practitioners: DA · DE · DS · ML · 8 x 45 min · open-source stack + AWS notes
One growing artifact: RetailPulse, a retail sales data product. The atomic unit of this track is the diff - one reviewable change that can touch code, schema, data expectations, and a model at once, and either passes every gate or does not. Every session adds one more thing a single safe change can now move. You build each capability by hand first, so when a machine writes the next one, you are the one who can read it.
Foundations: repo, Git flow, CI, Docker
Scaffold RetailPulse, wire pre-commit and a GitHub Actions pipeline that lints and tests every PR, and containerize it.
🌀 Builder 2 · easyOrchestration: pipelines as code
Turn a script into an Airflow DAG - tasks, schedules, idempotency, backfills - with a Dagster comparison and env config.
✅ Builder 3 · mediumData testing and contracts
Great Expectations and dbt tests, a data contract for the sales table, and a quality gate that fails the PR on breach.
🗄️ Builder 4 · mediumDBOps: schema migrations that never break prod
Version the database with Flyway/Alembic, run migrations in CI/CD, and rehearse the expand-contract rollback.
🔬 Builder 5 · hands-onMLOps I: experiment tracking and the registry
A demand-forecast model with MLflow tracking, a model registry, reproducible runs, and model CI that gates on a metric.
🚀 Builder 6 · hands-onMLOps II: deploy, monitor, retrain
Serve the model with FastAPI + Docker, watch for drift with Evidently, and wire the trigger that says retrain now.
📡 Builder 7 · hands-onObservability, secrets and IaC
Pipeline logs and metrics, freshness and volume alerts, secrets done right, and Docker Compose that stands the stack up.
🎓 Builder 8 · hardestFull CD to prod: the capstone
Promote dev -> staging -> prod, release data, model and database together safely, tag v1.0, and run the graduation checklist.
Choose your path 🗺️
Two doors, one course. Leaders never see a terminal; builders never sit through strategy they have not first felt break in a pipeline.
The knowledge map 🧠
Both tracks at a glance - hover a session to spotlight its concepts, click any node to jump in.