We organise your company’s information, prepare instructions for recurring tasks and test the setup with your people. Your team starts with approved material, follows an agreed structure and knows what to check before using the result.
Let’s talk
For the tasks we agree to organise, we prepare the material, instructions and review steps with your team.
The approved reference material, the expected structure and the instructions for preparing and checking the work.
The information for that particular task: this month’s figures, the client’s requirements or the campaign brief.
The facts, figures, terms and decisions that need human judgement before the work is used or sent.
The setup
We bring together what AI needs to know for the tasks we agree on: what you offer, who your customers are, which terms apply, and examples of the work you expect. We select up-to-date material and establish which source takes priority when different versions exist.
Each project brings together the information, instructions and conversations for a specific area of work. Your team can find what they need and pick up where they left off without explaining everything again. We define who uses each project, who keeps it up to date, and which information needs restricted access.
We write instructions for the structure and tone of the output, the questions AI should ask, and the checks people need to make. We agree on which data can be used and when approval is required before work is used or sent.
We work with your people on real tasks. We show them how to explain what they need, provide the right information, review the first draft and ask for specific changes. We keep reusable instructions for recurring tasks.
Everyday work
We choose the first tasks together. Here are a few examples.
Example
Approved monthly figures, metric definitions and the previous report.
A first draft of commentary on changes, with questions that need further review.
The figures, the reasons for changes and the conclusions drawn for management.
What we measure Preparation time and corrections needed.
Example
The role description, company procedures and approved onboarding material.
An onboarding guide tailored to the role.
Completeness, whether procedures are current, and whether the information is appropriate for the new employee to access.
What we measure Drafting time and omissions.
Example
The customer’s request, company policies and approved responses.
A draft reply, highlighting any missing information.
Accuracy, exceptions and any commitments made before the reply is sent.
What we measure Response time and rounds of revision.
Example
Client notes, service information, prices and terms.
A draft proposal following the agreed structure.
Whether the offer fits the client’s needs, and whether prices and terms are correct.
What we measure Drafting time and completeness.
Example
A short campaign brief, target audience, approved product information, offer details and examples of brand tone.
Draft campaign emails and social posts, with a consistent message adapted to each channel.
Product details and claims, offer terms, tone and suitability for each channel before publication.
What we measure Preparation time, rounds of revision and consistency across channels.
Scheduled tasks
Once the information and projects are organised, we identify which tasks are worth scheduling. We agree on what needs to be checked, which sources to use and what you want to receive.
Every Monday: An update on developments at selected competitors, with sources and dates.
Before the weekly meeting: A summary of the team’s updated action tracker, including deadlines and items that need a decision.
Every morning: A summary of requests and outstanding items from agreed email or calendar sources, where the required access is supported.
Before agreeing on each task, we check the tool’s capabilities, subscription plan and available sources. We clarify whether it needs an additional connection or a file that your team keeps up to date.
Delivery and implementation
We start with a conversation about your team, its current tasks and how you use AI today. We agree on the scope and request the files we need. We set up the environment, test it with the people who will use it and make any necessary changes before handover.
Your company provides a coordinator, the people who know the selected tasks and the approved material we will use. The proposal sets out the time required from your team and the deliverables.
Results and measurement
We aim to reduce the time spent repeating instructions and searching for information, improve consistency and cut down on revisions. Your team learns how to ask for the work they need and check what AI produces.
For selected tasks, we compare completion time and revisions before and after the setup. We also check whether people are using what we have built and where they still need help.
Once your team is regularly using the organised setup, we look at processes with more steps and connections to other systems. For example, how a new request becomes a draft response, goes through approval and is recorded in the right system.
Workflows and agents are designed around needs identified through day-to-day use, with a separate scope and proposal.
Co-founder & CEO
Theofilos Vasileiadis is a serial entrepreneur, bestselling author, and co-founder and Chairman of Kariera Group. He works with companies on growth, technology, and the practical adoption of AI. He is the founder of Flow22 and Stackworth22, which helps companies unlock the value of their software assets.
LinkedIn
AI Consultant
Olga Mavi is a product leader and entrepreneur with experience building and scaling technology products. She co-founded Socital and served as VP of Product at Kariera Group, leading its international HR technology portfolio. At Flow22, she brings her product and technology experience to help companies put AI to work.
LinkedIn
AI Consultant
Georgina Tseroni is a partnerships and talent professional with experience in technology recruitment and professional development. She has worked with Workearly and Kariera Group, connecting companies and people with opportunities to grow. At Flow22, she works with the team to help businesses bring AI into their everyday work.
LinkedInBefore we start
We organise the information, instructions and working practices around your team’s tasks. People start with shared information and clear criteria, helping reduce repeated explanations and corrections.
We choose the right tool for the agreed scope. If your company already uses one, we first look at how to organise it more effectively.
We agree on which material is needed, who can access it and what stays out. We review the selected tool’s subscription plan and settings before using company information.
We need a coordinator, approved material and users who can test the setup on real tasks. We agree on the meetings and time required before we start.
We select suitable tasks after checking the sources, access and tool capabilities. Where someone needs to update information or approve an output, we include that step in the process.
The proposal sets out the scope, deliverables, timeline and fee. Tool subscriptions and any additional connections are listed separately.
We agree on a person responsible within your team and provide instructions for updating the material. Ongoing support or further development can be agreed separately.
The first step
Tell us which tool you use today, which tasks keep coming up and where time gets lost. We’ll discuss what needs to be set up and propose a specific scope of work.
