Start from a role or one recurring task.
Use the role chips or task box to choose a real workflow your team repeats, not a broad promise to “use AI more”.
Decide where AI could help · Should AI do it?
Decide what you can hand to AI, and what stays yours.
Tap your role or describe a task. You get what AI can take, what you keep, and a handover plan for this week.
Choosing for a team? Use this as the team handover agenda, not a course or a new portal.
Decision rule
Hand over the repeatable part. Keep the consequence.
If the work affects money, safety, law, trust, private data, or a person's next step, use AI to prepare a draft or help you check the work. Keep the decision with a person.
Where AI fits for data analysts who need the numbers to hold up
Each row is a reviewed workflow: what AI takes, what you keep, and the tool that fits.
Shares the public view and role only, without your typed task.
Understand a file or spreadsheet
AI takes- Reading files and checking columns and data types
- Code backed analysis and formula generation
- Cross file synthesis and editable outputs
You keepThe person decides what the data means, whether it is fit for the stated purpose and which action the evidence justifies. The tool can calculate quickly, but it does not own the business definition or the consequence of a misleading metric.
Research a topic with reliable sources
AI takes- Editable research planning
- Web, file and connected source retrieval
- A summary with links to its sources
You keepThe person decides whether the evidence is credible, relevant and sufficient for the real decision. AI can assemble and challenge a case, but it cannot set the organisation's evidence threshold or accept the consequence of being wrong.
Compare difficult options
AI takes- Criteria generation and structured comparison
- Scenario and sensitivity analysis
- Reasons against each option and ways it could fail
You keepThe person or authorised group owns the criteria, the weighting, the acceptable trade off and the final commitment. AI can improve the reasoning process but cannot supply organisational legitimacy or bear the outcome.
Create a presentation from source material
AI takes- Multi file synthesis and storyline drafting
- Editable presentation and document creation
- Charting, visual concepts and critique
You keepThe presenter owns the argument, the ethical use of evidence, the final ask and the delivery. AI can accelerate synthesis and production, but it cannot read the room or take responsibility for persuading people with a misleading story.
What stays yours
Understand a file or spreadsheet
Bring these before you hand anything over.
- Define what the numbers mean
- You know what each field, status and measure means in the business, which definition is current and which exceptions must be handled deliberately.
- Explain where the data came from
- You know how the data was collected, transformed and refreshed, where manual behaviour enters, and which changes in process can create a false trend.
- Domain interpretation
- You can judge whether the pattern makes sense in the real world, identify confounding events and translate a statistical result into an appropriate operational conclusion.
Your handover plan
Your first-week plan
for data analysts who need the numbers to hold up
4 workflows · about 80 min of setup · a small team can put a name against each one.
Stays with a person, always (4)
- The main failure is a technically clean answer to the wrong data question. Errors in definitions, units, coverage or source process can survive every calculation and become more persuasive once they are charted.
- The main failure is false confidence through volume. A long, cited report can hide a poor research question, repeated secondary sources, missing evidence and conclusions that travel further than the sources allow.
- The main failure is outsourced judgement disguised as objectivity. A neat matrix can conceal arbitrary weights, missing stakeholders and assumptions that make the preferred option appear inevitable.
- The main failure is persuasive polish without earned meaning. A generated deck can look board ready while carrying the wrong denominator, a generic storyline or an ask that ignores the relationships and history in the room.
First move to paste into ChatGPT
Before analysing this file, tell me what each field appears to mean, identify missing or inconsistent data and ask about ambiguous columns. Then answer: [question]. Show the calculation method, call out assumptions, keep source rows traceable and give me one chart only if it makes the result easier to verify.
First move to paste into ChatGPT
Research [topic] to help me decide [decision]. Focus on [country/industry] and information from [time period]. Prioritise primary sources and credible independent analysis. Compare the strongest options, disagreements and unknowns. Link every material claim. Finish with what the evidence supports, what it does not support and five questions I should investigate next.
First move to paste into ChatGPT
Help me think through this decision without choosing for me: [decision]. Options: [list]. My constraints: [list]. First propose a short set of decision criteria and ask me to correct them. Then compare the options, show the main trade-offs, mark unverified facts and make the strongest case against whichever option currently looks best.
First move to paste into ChatGPT
Use these files to propose a [length]-slide presentation for [audience]. The outcome I need is [decision or action]. Start by showing only the storyline, the job of each slide and the source material each slide would use. Keep source facts traceable and mark missing evidence. Wait for my approval before drafting the full presentation.
Start here this week
Understand a file or spreadsheet in ChatGPT. Paste its first move, then check the result against the after-you-hand-off list before anyone relies on it.
For a small team
Use the same triage before you buy training or another tool.
AI Fit is not a course or a corporate portal. It is a way to turn the reviewed role and workflow records into one grounded team conversation: what can be handed over, who checks it, and where a person still owns the call.
Put a name beside each handover step.
The plan already includes owner slots. A team lead can decide who tries the first move and who checks the result before anyone relies on it.
Use the Before, While, After and Never checks as your agenda.
Agree what information to provide, what to supervise, what to verify afterwards and which work must stay with a person.
Reusable agenda
Copy the public team review steps.
This is the exact agenda the button copies. It deliberately leaves out typed task text, private details, customer data and internal names.
- Pick one role or recurring task the team actually repeats.
- Read the handover split: what AI can take and what a person still owns.
- Choose one workflow to try this week and put a named owner beside it.
- Use the Before / While / After / Never checks as the review agenda.
- Ask Bit only to tailor the first-week plan after the team agrees the boundary.
Do not add typed task text, private details, customer data or internal names to the shared agenda.
Your next step
Tailor the plan for the real week in front of you.
Ask Bit to adapt this plan to your task, time and constraints.
