AI for data and analytics

AI training programme · 6 modules · 3 Skilledge Hyper labs

"Ask your data in plain English" works far better when the data has been defined properly first.

In Snowflake's own test, the same model answered 57% of benchmark questions correctly as raw text-to-SQL and 78% once it had a semantic model to work from. It's a vendor number, but the lesson holds: the definitions matter more than the model. So this programme spends real time on the dull part, which revenue figure is the revenue figure, then gets to the fun part.

At a glance

Who it’s for
Data and BI analysts, analytics engineers, business users who self-serve reports.
Tracks
Data & BI

Modules

  1. Text-to-SQLGetting good queries, and catching the ones that run fine and return the wrong number.
  2. Semantic layersWriting metric definitions an AI can't misread.
  3. Python analysis with AICleaning, joining and exploring data in notebooks, with every step explained.
  4. Data qualityUsing AI to profile data, find anomalies and draft validation rules.
  5. AI-built dashboardsReviewing what the assistant made before a leader sees it.
  6. Telling the storyTurning an analysis into a recommendation someone can act on.

Skilledge Hyper lab exercises

Lab designs: what a cohort does in Skilledge Hyper, on mock data.

  • Twenty questionsA mock sales warehouse and twenty business questions, answered with AI-written SQL. Hyper grades each answer against the true result.
  • Fix the definitionsAdd a semantic layer, rerun the same twenty questions and compare accuracy.
  • The misleading chartAn AI-built dashboard with three quiet errors in it. Find them.

Tools covered

AI assistants in Power BI, Tableau, Databricks and Snowflake; general-purpose models for SQL and Python; your own warehouse where access allows.

Sources

Related programmes

Want this for your team? Tell us who it’s for and what they need to do differently after it.