DecisionBox for business: videos for people who act on findings
"DecisionBox for business" is a free video series for heads of marketing, sales, product, operations and finance who use DecisionBox findings to make decisions. Each episode teaches one task in a few minutes, using the real product on a public e-commerce demo dataset. Seven episodes are live.
Who the series is for
Your data team connects DecisionBox to the warehouse. These videos start after that, with reading findings and deciding which recommendations to act on. They also show how to tell DecisionBox what matters to your business. You don't need to write SQL or set anything up to follow them.
The presenter is an AI-generated host, and every video carries a label saying so.
What episode 1 covers
The first episode walks through a finding's numbers and the verdict from the two AI agents in DecisionBox that check it. It also shows how to open the queries behind the finding, so you can decide how much weight to give it.
Episode 1 on YouTube: How to read an AI finding, and how much to trust it
The validation page and our post on how DecisionBox checks a finding cover the same checks in writing.
The episodes so far
| # | Episode | What you'll learn | Edition |
|---|---|---|---|
| 1 | How to read an AI finding, and how much to trust it | Read a finding's numbers and validation verdict. Open the queries behind it and give feedback. | Community edition |
| 2 | Teach DecisionBox your business definitions | See what DecisionBox learns about your tables on its own and correct a description in the semantic layer. Pin business definitions as notes and see Ask use them. | Table descriptions: Community edition; notes and Ask: Enterprise |
| 3 | Ask your data a question in plain English | Ask a question and read the answer with its chart. Open the query or ask a follow-up. | Enterprise |
| 4 | From finding to decision: how to read an AI recommendation | Read a recommendation's expected impact and steps. Follow it back to its findings and check its validation result before acting. | Community edition |
| 7 | Put AI data discovery on a schedule | Create a weekly schedule and check its next and last run. Read what a scheduled run found. | Enterprise |
| 8 | Steer an AI analyst: feedback, playbook and targets | Vote on findings and say why. Edit the playbook DecisionBox generated from your data. | Feedback and playbook editing: Community edition; generating a playbook: Enterprise |
| 9 | Discovery Ledger: track AI findings across runs | See what earlier runs covered and what is still open. Check what each run picked up. | Enterprise |
The whole set is in the DecisionBox for business playlist. The gaps in the numbering are episodes still in production; the playlist fills in as they go live.
Which edition covers what
The open-source Community edition includes findings and their checks, recommendations, table descriptions, feedback and the playbook editor.
Asking new questions of your live warehouse data, scheduled runs, notes and the Discovery Ledger are in Enterprise. Generating a playbook from your data is in Enterprise too.
On your own data
The public demo dataset lets you follow the videos without access to private data. For your data team's setup work, there's the first discovery guide.
