Coming Soon workshop
When Wrong Numbers Look Right: Validating AI Analysis
AI
Data Prep
Critical Thinking
Notify me when registration is open
AI can produce an analysis in seconds, and most of the time it looks completely reasonable. But a wrong number and a right number look identical on screen, and the mistakes that slip through are rarely arithmetic. They come from a filter nobody mentioned, a join that quietly duplicated rows, or a metric definition that was never agreed, and checking every one of them manually would defeat the point of using AI in the first place.
In this session, you'll validate data analysis outputs using AI step by step, with a simple prompt template that keeps you in control of what gets tested. You'll write down what you expect and why, have AI turn that into short Python or SQL you can actually read, and run the checks. You'll leave with the template, worked examples, and a set of prompts you can point at your own data straight away.
What you'll learn
Apply structured prompt templates to check AI-generated work
Stay in control of what gets tested instead of trusting AI outputs at face value.
Identify the errors that survive a casual review
Silently dropped rows, bad joins, and ambiguous metrics are the mistakes that reach stakeholders, but rarely look wrong on screen.
Direct AI to write short, readable Python or SQL that validates a result
You can run and understand these checks even if you don't write code yourself.
Build a reusable set of validation prompts you can apply to any dataset
The same prompts work across datasets and AI tools, so the process carries over to your own work.
Who should join this workshop?
Analysts who already use AI day to day and want a reliable way to check output before it reaches a stakeholder
Non-technical professionals who work with data but don't code and want to verify their numbers with AI
Data professionals asked to sign off on AI-assisted work who need a process they can stand behind
Team leads who want a shared standard their team can apply to AI-generated analysis
Prerequisites
Access to an AI tool like Claude, ChatGPT, Copilot or Gemini(paid version recommended)
Mo Chen
Data & AI Educator, Data with Mo
Mo Chen is a data and analytics professional with experience across finance, risk, investment banking, and business insights. After earning his MSc in Finance & Economics from the University of St Andrews in 2018, he moved into investment banking as a Data & Analytics Manager. Since 2023, Mo has grown a major online presence creating data and career content, reaching 200K+ YouTube subscribers, 8M+ views, and 50K+ LinkedIn followers.
