// about

The hands-on half of cert prep.

labrun is a hands-on companion to whatever video course got you this far. Real AWS, Azure, and Databricks accounts. Real Python notebooks. Live API checks on every checkpoint. No simulations, no multiple choice, no honor system.

// what labrun is

A lab platform, not a course.

Each lab is a Jupyter notebook you open in Google Colab on a free runtime. Cells have intentional blanks. You fill them in (resource names, parameters, transformations) and run the cell. A companion panel calls the live cloud API against the resources you just created.

If the resource matches the lab spec, the checkpoint goes green. If it does not, the panel returns a one-sentence directional nudge without spoiling the answer. If you are stuck, a 3-tier hint system steps you up: nudge first, pointer second, full answer third.

When the lab ends, automatic cleanup tears every resource down, verified against the same APIs that ran the checkpoints. Labs are designed to cap under $2 per session; most learners spend $5 to $15 total across a full cert path.

// certifications

6 active certs, 71 hands-on labs.

Every lab is built around the live exam blueprint and reauthored when the provider ships a new exam version.

  • AWS · DEA-C01

    Data Engineer Associate

    12 labs · ~13h · Intermediate

  • AWS · SAA-C03

    Solutions Architect Associate

    12 labs · ~11h · Intermediate

  • Databricks

    Data Engineer Associate

    12 labs · ~12h · Intermediate

  • AWS · MLA-C01

    Machine Learning Engineer Associate

    12 labs · ~13h · Intermediate

  • Azure · AI-103

    AI Apps and Agents Developer Associate

    11 labs · ~11h · Intermediate

  • Databricks

    Generative AI Engineer Associate

    12 labs · ~13h · Advanced

On the roadmap

  • AWS AI Practitioner (AIF-C01) · Q4 2026
  • Snowflake SnowPro Core · Q1 2027
  • dbt Analytics Engineer · Q1 2027
  • HashiCorp Terraform Associate (003) · Q2 2027
  • Snowflake Advanced Data Engineer · Q3 2027

// job prep tracks

Role-based tracks for the senior loop.

Sequenced labs that build a portfolio of shipped systems, not chatbot demos. Separate plan targeting $49.99/mo (TBD).

Data Engineering (16 labs)

Stand up production pipelines, model the warehouse, debug silent failures, and walk into a senior DE loop with the receipts.

  • 5 labs · ~7h

    Pipeline Engineering

    Build, orchestrate, backfill, and debug production pipelines under real failure modes.

  • 5 labs · ~7h

    Warehouse & Lakehouse Mastery

    Model the data, pick the right table format, tune for cost and speed.

  • 3 labs · ~4h

    Streaming & Change Data Capture

    Event-driven ingestion, CDC, schema evolution, and exactly-once delivery.

  • 3 labs · ~4h

    Data Reliability Engineering

    Quality gates, on-call triage, and cost monitoring for production data systems.

AI/ML Engineering (20 labs)

Ship production AI systems with eval, observability, and cost discipline. Walk into a senior AI loop with shipped systems, not chatbot demos.

  • 5 labs · ~8h

    Retrieval Systems

    Hybrid retrieval, reranking, multimodal RAG, and the eval harness that catches a bad deploy.

  • 5 labs · ~9h

    LLM Applications & Agents

    Structured output, tool calling, durable agents, and the fine-tune decision.

  • 5 labs · ~9h

    Inference Serving & Optimization

    Self-hosted inference, quantization, A/B traffic, and multi-provider gateway routing.

  • 5 labs · ~7h

    Eval, Observability & Safety

    The moat for production AI: regression evals, traces, prompt-injection defense, and drift detection.

// how labrun compares

Where labrun fits.

Each option below is good at a different thing. labrun is specifically the hands-on validation layer; use it alongside a video course, not instead of one.

OptionFormatTypical priceReal cloud validation
labrunHands-on notebooks with live API checks$24.99/moYes, every checkpoint
A Cloud Guru / PluralsightVideo + sandbox labs$30 to $50/moPartial (sandbox accounts)
WhizlabsVideo + practice exams + sim labs$20 to $40/moNo (simulated)
Stephane Maarek (Udemy)Video lectures$15 to $20 lifetimeNo (video only)
Adrian CantrillVideo + manual hands-on~$40 per certManual (no live validation)
Tutorials DojoPractice exams + study notes$15 to $30 per certNo
AWS Skill BuilderFirst-party video + Cloud Quest$29/moPartial (CloudQuest)

// questions

Frequently asked.

>What is labrun?
labrun is a hands-on cert prep platform. Students open a Jupyter notebook in Google Colab, fill in intentional blanks against their own AWS, Azure, or Databricks account, and a companion panel calls the live cloud API to validate every checkpoint. There are no simulations, multiple choice, or honor-system tasks. Automatic cleanup tears every resource down at the end so the cloud bill stays near zero.
>How is labrun different from A Cloud Guru, Whizlabs, Stephane Maarek, or Adrian Cantrill?
Those are video courses (or sandbox simulators) that explain the concept. labrun is the hands-on validation layer: after the video explains what a Glue ETL job is, a labrun lab makes you build one in your own AWS account and validates it against the live AWS API. Most learners use labrun alongside one of the video courses, not instead of one.
>How much does labrun cost?
The cert plan is $24.99 per month (monthly or 3-month plan). It includes full access to every cert and every lab labrun ships. New certs are added with no upgrade required. The job prep tracks plan launches separately, targeting $49.99 per month (TBD).
>Do I need my own AWS, Azure, or Databricks account?
Yes. Labs run in your own cloud account so you own everything you build and the validation calls hit the real provider APIs. Each lab documents the IAM user, managed policies, and credentials it needs. Labs are designed to cap under $2 per session; most learners spend $5 to $15 total across a full cert path.
>Which certifications does labrun cover today?
Six active certifications, 71 labs total. AWS Data Engineer Associate (DEA-C01), AWS Solutions Architect Associate (SAA-C03), AWS Machine Learning Engineer Associate (MLA-C01), Databricks Data Engineer Associate, Azure AI Apps and Agents Developer Associate (AI-103), and Databricks Generative AI Engineer Associate. AWS AI Practitioner, SnowPro Core, dbt Analytics Engineer, Terraform Associate, and Snowflake Advanced DE are on the 2026 to 2027 roadmap.
>What are the job prep tracks?
Role-based prep tracks for engineers heading into senior interview loops. Data Engineering (16 labs across 4 sub-tracks: Pipeline Engineering, Warehouse & Lakehouse Mastery, Streaming & CDC, Data Reliability) and AI/ML Engineering (20 labs across 4 sub-tracks: Retrieval Systems, LLM Applications & Agents, Inference Serving & Optimization, Eval & Observability & Safety) are live. DevOps & Platform Engineering is coming Q4 2026.
>What happens if I cancel?
Access continues through the end of your billing period. The notebooks themselves live in your own Google Drive, so the work survives cancellation.
>Who built labrun?
Taqi Moghal. Data engineer; solo founder. labrun was built because cert prep that ends at the practice exam leaves you sitting in an interview unable to actually build the system the exam was about. Reach Taqi at taqi@labrun.dev.

// who built this

Taqi.

Data engineer; solo founder. He got tired of cert prep that ends at the practice exam: pass the test, then sit in interviews unable to actually build the thing you just certified on. labrun is the prep tool he wishes he had a year ago. Real cloud, real resources, real checkpoint validation.

Want to talk shop? Email taqi@labrun.dev.