AIStrategy.gr Library

Archives: Playbooks

The first 90 days — an AI adoption working plan

A sequencing guide for a bounded organisational initiative. These time windows organise the work; they are not a promise of delivery. Days 1–15: define and baseline Choose one operational problem, name its owner and speak to the people doing the work. Record volume, quality and effort. Identify the affected users and the decisions the system […]

Publish an AI case — an evidence submission template

Use this format to make a case useful to another organisation. Unavailable information should be marked unavailable rather than estimated without a basis. Context and baseline Describe the organisation, task, users, period and starting process. Identify the problem and the reason the existing approach was insufficient. Include the baseline measure and how it was collected. […]

AI supplier review — questions and required evidence

A discussion pack for procurement, technology, security and the operational owner. The actual contract and use require appropriate professional review. Product and workflow Which exact task does the product support? What does it refuse to do? Which actions need approval? Ask for examples of known failures and the conditions under which the supplier recommends human […]

Data readiness — a workshop worksheet

Use this worksheet with the data owner, operational team and security representative before selecting a model. Inventory the inputs For each data source record owner, purpose, format, volume, refresh frequency and access route. Identify whether it is authoritative, copied or manually assembled. Note missing periods and known changes in collection practices. Check identity and time […]

Retrieval assistant — an acceptance test pack

A grounded assistant needs tests for retrieval, answer quality and the action a user takes afterwards. Prepare representative questions Create questions from real work. Include straightforward requests, ambiguous questions, outdated information, requests outside scope and examples with no answer in the approved corpus. Keep a held-out set that the builders do not use to tune […]

AI pilot scorecard — a decision template

Use one scorecard from problem selection to the decision to scale. Record observations separately from assumptions. 1. Define the service Write: task; user; accountable owner; volume; current handling time; current error rate; cost of failure; and the person authorised to accept the result. If the owner cannot describe the existing process, establish that baseline first. […]

Move from pilot to production

A practical working guide from AIStrategy.gr. Define the service owner Assign accountability for performance, incidents, access and operating cost. Identify the team responsible after the pilot ends. Validate in real conditions Evaluate representative workloads, edge cases, user behaviour and fallback procedures. Keep the baseline and evaluation set versioned. Monitor and respond Track quality, drift, costs […]

Buy AI responsibly

A practical working guide from AIStrategy.gr. Describe the workflow Give suppliers the task, data boundaries, expected users and acceptance criteria. Identify who approves outputs and handles incidents. Ask for evidence Request evaluation results, security documentation, data-use terms, retention controls, subprocessors and known limitations. Plan the exit Ask how data and configuration can be exported, how […]

Build the AI business case

A practical working guide from AIStrategy.gr. Measure the baseline Record handling time, error rate, cost per task and volume. Separate observed measurements from assumptions. Calculate total cost Include integration, model use, data preparation, review, security, training, operations and the cost of switching suppliers. Test the benefit Compare pilot results against the baseline. Check whether time […]

Choose the first use case

A practical working guide from AIStrategy.gr. Define the decision or task Describe one task, who performs it, how often it occurs and what failure costs. Avoid starting with a model or product name. Compare value and feasibility Score candidate tasks for measurable benefit, data access, integration effort and risk. Prefer a bounded workflow with an […]