We help your people use AI in their everyday work
Where AI is already in use, often only a fraction of its capabilities are used. People know individual tools but do not know how to bring them into their work or combine them into one workflow. With AI Buddy, we guide employees through their real work tasks so they can use AI on their own. We document the change in their work by measuring it.
Leaders in automotive, banking, and telecommunications rely on our solutions
We work with your real tasks
Not practice examples, but the meetings, reports, and tasks ahead of you.
Your employee does the work; we guide them
The guide suggests an approach and explains it; they do not do the task for the employee.
We measure the change in work
We measure the baseline and the difference on specific activities.
Situations where AI is not used to its full extent
These are the situations we encounter most often.
Common approaches and their limits
| Approach | Limit |
|---|---|
| Doing nothing | AI is used without shared rules, and the benefit cannot be evaluated |
| Training only | improves knowledge but may not change day-to-day work |
| Buying licenses only | without rules, content, and guidance, the tool falls out of use after the first problems |
| In-house | possible and the right long-term path; in the short term, capacity is often lacking |
We do not question introducing AI in-house. Working with us helps prepare the people and practices the company then continues with on its own. Nor do we claim that training does not work. Training tailored to specific people serves its purpose, but it addresses something else: it teaches how to use a tool, while AI Buddy changes how work is done.
How it works
We guide people in working with AI directly on their work tasks
A guide who works with AI themselves starts by asking which tasks the employee is currently working on, and the work happens directly on those. The employee carries out the task; the guide suggests an approach, corrects, and explains the reasons. If the guide did the task for the employee, the employee would not learn the approach and would need help again with the next similar task. The guide also helps bring in the company's tools, systems, and materials the employee has access to.
We map how work actually happens in the area
We start from records of completed work in the systems you already use, not from questionnaires. The named employee's work is mapped; others' work only in aggregate, with their consent, and without sensitive data. The result is a picture of the work the company has not had before. The map is a means, not the goal.
We sort the activities
We determine which activities can be automated, where AI should help a person, and which should stay with people only, always with a justification. The company sees where it is worth investing. Activities suitable for automation are named and handed over; implementing them is a separate project. Flows in the organization →
We measure the baseline
On selected activities, we measure time, the number of manual steps, and how often work comes back for correction. Without this measurement, the benefit cannot be documented, so we do not start an area without it.
We work regularly on real tasks
Short meetings on the work employees are doing right now, without dedicated training days. The goal is for them to use the approaches on their own.
Procedures and AI assistants that work without us
We consider them handed over only when someone other than their author uses them. They stay in the company.
We measure the difference
On the same activities as at the start. The result is a quantified difference.
Mapping the work runs in parallel with preparing the rules for using AI, the list of approved tools, and the work environment.
Who takes part
Employees have support with their first tasks with AI. Our guide works with the named employee on their tasks. Colleagues we train help others directly in their work and stay in the company after the engagement ends. In the tool the employee works in, there is also an assistant prepared for the specific activity, with its procedure and templates. Before work with people begins, employees receive the data rules and the list of approved tools.
| Who | Side | Role |
|---|---|---|
| Guide | Aspecta | leads the joint work on the employee's tasks |
| Named employee | company | their work is improved and measured |
| Trained colleagues | company | help others and stay after the engagement ends |
| Assistant in the tool | tool | supports a specific activity according to its procedure |
| Area lead | company | provides the conditions, confirms the choice of activities, and signs off the results |
| Environment administrator | company | runs the work environment after handover |
| Others in the area | company | contribute briefly to mapping the work |
We show how to use AI even where it would not occur to you
The benefit is not only in answering the question asked, but also in what can be done and how it can be connected.
| Today | With AI |
|---|---|
| a report put together by hand every month | a procedure that builds it from the same sources |
| a large volume of documents nobody has time to read | analysis of all documents, not just a sample |
| information has to be obtained from a colleague | a step that finds it in materials the employee has access to |
| several tools and copying data between them by hand | one workflow in which the steps follow on from each other |
It is not a missing skill but a missing picture of what is possible. A standard course does not provide that.
We introduce AI Buddy area by area
Each area has a named employee whose work is measured and whom the joint work concerns. Other colleagues contribute briefly to mapping the work, and we train selected people to help others. We open an area where its lead is interested; often this is management and its own agenda. The scope of the engagement is determined by the number of areas involved. In each, we keep the same way of measuring so the results can be evaluated.
What we measure and what we do not
We measure two things. The first is inputs: what the employee did not have before the start and has afterward, by an agreed date. The second is outputs: specific activities, their time, the number of manual steps, and how often work comes back for correction; depending on the area, also other indicators, such as the share of meetings with minutes or the share of requests returned for clarification. If outputs prepared with AI are repeatedly reworked, it signals the activity was poorly chosen, and we address it. We do not measure the number of prompts, how often AI is used, completed training, or levels reached: these are not results and do not count toward sign-off.
Protecting employees is part of the contract
We measure activities, not people. For teams, we measure only in aggregate; for named employees, their own work, with their consent. No output of the engagement is used for decisions about headcount or about a specific person. Individual productivity measurement is not part of the engagement in any form. Before an area starts, the engagement is discussed with employee representatives. These conditions protect employees and build the trust without which measurement would make no sense. Monitoring an individual's performance through AI also conflicts with employee privacy, so we do not offer it.
How we support long-term use of AI
Introducing AI most often fails in two places, and AI Buddy addresses both by design. The first is the first experience: if an employee tries AI on an unsuitable task or does not know what they may use, there is usually no second attempt. That is why we do not start with a tool but with rules and the list of approved tools, and the first tasks are guided by someone who already works with AI on this kind of work. The second is missing evidence: the baseline is often not measured at all, or only afterward. With AI Buddy, we measure it before work with people begins.
From individual work to business processes
A procedure created in one employee's work can be taken up by colleagues too. At that point it needs a designated owner and rules, and it becomes part of a business process. Over time, some activities may be taken over entirely by an automated process. That is a natural development; the employee can focus more on oversight, decisions, and tasks that require human judgment. The work map names candidates for such changes; implementing them is a separate project. Flows in the organization →
How AI Buddy differs
We measure work, not usage
Growth in usage is activity, not a result. We measure the change in specific activities and quantify the difference, whatever it turns out to be. We state just as clearly what we do not measure.
The baseline is a condition
We measure it before the first meeting; without it, we do not start an area.
Every output has an agreed acceptance criterion
A procedure counts as handed over when someone other than the author uses it. The environment, when your administrator carries out an operational task without us.
We work on current tasks
Not on model situations or on dedicated days. It takes more thorough preparation, but it makes it possible to verify whether employees can use the approaches on their own.
Environment, rules, and content are part of the engagement
We do not leave them to someone else, because that is where first attempts most often fail. For each item, it is stated which activity it enables.
Exclusions are agreed in advance
What is not in scope is stated before the start, not at sign-off.
We do not promise what the environment cannot support
Where a rule can be enforced technically, we enforce it and verify it with a test. Where it cannot, we say so openly.
When another solution may be a better fit
If you need a fast first visible result, a lower price with the option of subsidies, a well-known international name, or a simple program with a fixed length, other offers may suit you better. A course and AI Buddy do not compete; they address different needs.
What you get
For every output, we define in advance how we will verify that it is complete. The work environment, the content, and the results of working with people are evaluated and signed off separately.
Employees who work with AI on their own
Thanks to support with their first tasks, they stick with it.
Colleagues who master the approaches and teach others
Further areas can then be led by someone from your company.
A measured baseline and a quantified difference
on the same activities.
Procedures and AI assistants that work without us
and are used by someone other than their author.
Data rules and a list of approved tools
One binding document employees can easily find, and a procedure for requesting a new tool, with a responsible person and a deadline. Data →
A work environment with managed access
and a separate regime for sensitive data. It is part of the preparation; whether you run it yourselves after the engagement or we run it is a separate decision. Tools for work →
Verified content about the company that assistants answer from
, with a traceable source and validity as of a date. Analyses, commercial materials, and information about specific people are not part of it.
Use of AI according to shared company rules
Verified by employees being able to apply the rule in their work.
A map of the real work in the area and a list of candidates for further improvements
, with justification. It is a by-product of the engagement and input for decisions on further investment. Flows in the organization →
How we verify the results of working with AI
We do not promise a specific result in advance. We commit to giving you a measured number at the end and to showing how it was produced; we do not promise its value in advance.
| When | What can be documented |
|---|---|
| after the first meetings | the first procedure the employee used on their own, which colleagues can also take on |
| after the first meetings | the first activity that was missing from the work map because nobody considered it solvable |
| at the first evaluation point | the first quantified difference on a specific activity |
| at the end | the difference on all selected activities, and procedures used by someone other than the author |
We agree on evaluation points before an area starts. Time is counted from the start of the area, not from signing the contract; individual areas do not have to start at the same time.
What we need from you
These are the conditions of the engagement, and it is good to know them before signing.
A named employee in the area
whose work is measured. Their time commitment is on the order of a few days over the whole period.
An area lead
who provides the conditions, confirms the choice of activities, and signs off the results. On the order of a few days per month.
Colleagues we train
to help others. On the order of a few hours per week.
Specific activities to measure
You do not need a baseline prepared; we measure it.
Access to records of work in the area's systems
If they are not available, the map is based on what employees say about their work, and we state this openly.
Discussion with employee representatives
before the area starts.
A budget for running the work environment after the engagement
Without it, we do not switch the environment on.
From employees in the area, we only need a short answer with each output: whether they will use it and whether they could have done it without AI. No reports, dedicated days, or training blocks.
For each engagement, we state the assumptions its scope is based on and check them against real data at the start. If they do not hold, we say so immediately, and the engagement can be stopped. If the company cannot meet these conditions yet, we choose a different scope or postpone the start.
What AI Buddy does not cover
Training, an audit, or education in the sense of subsidy schemes
The subject is changing the work in a specific role; the work map is a means, not the goal. If you need a course on operating a tool, training is a better fit. Under certain conditions, part of the engagement may be supported from public funds; that is not part of the offer, but we can advise where to look for information.
Company-wide instruction for all employees on using AI
That is a separate activity. AI Buddy builds on it and focuses on specific roles.
Assessing employee performance
In no form; no output is used for decisions about a specific person.
Process automation or connecting systems
We name candidates for automation and hand them over; implementing them is a separate project with its own scope and responsibility. Flows in the organization →
Your own environment for data that must not leave the company, or running the work environment after the engagement
These are separate decisions with their own budget. Infrastructure →
Technical enforcement of rules where the environment does not allow it, or an internal team that rolls out AI across the whole company
In the first case, we provide organizational measures and point this out in advance. If AI Buddy eventually becomes a way of working the company runs itself, that is an additional outcome, not a commitment.
Frequently asked questions
How does AI Buddy differ from standard AI training?
Training teaches you how to use a specific tool. With AI Buddy, employees work with AI directly on their day-to-day tasks, guided by an experienced specialist. The goal is for them to use the new approaches on their own. We verify the difference on specific activities: we measure the baseline and then quantify what has changed in the work.
How much time do we need to free up for employees?
AI Buddy runs alongside regular work. Employees work on tasks they would do anyway and do not need to set aside whole days for training. The engagement does, however, require their active involvement. For the employee whose work we measure, it is on the order of days over the whole engagement. We agree on the required time and level of involvement in advance.
How do you protect our business data?
Before work begins, we set rules that define which data may be used in which AI tools. Employees have a list of approved tools and clear guidance on what they may enter. Where the technical environment allows it, compliance with the rules is also checked. Where technical control is not possible, we say so openly. Data that must not leave a defined environment is processed in a separate regime. Data →
What is the difference between AI Buddy and AIGrip?
AI Buddy is a service in which we help employees bring AI into their work and measure the change achieved. AIGrip is a framework for working with AI that helps keep context, decisions, and procedures intact in longer and more complex tasks. They can complement each other but do not depend on each other. We can deliver AI Buddy without AIGrip, and you can use AIGrip on its own. Framework for work →
What will we have after the engagement ends?
Employees who can use AI in their tasks on their own, together with the procedures, AI assistants, rules, and working materials created. The agreed scope also includes the list of approved tools, prepared company content, and the work environment. We hand over outputs in an open format so your people can keep using and developing them. We verify their independence in practice, for example by their ability to add content without our help. You can run the environment yourselves or entrust it to us.
Why work with Aspecta?
We use the approaches we introduce at customers in our own work. We have established rules for using AI, a list of approved tools, and measurement of changes in specific activities. We already deliver AI Buddy to a research organization and are introducing it in an enterprise environment. We do not promise specific savings or a level of improvement in advance. But we provide measurement that shows the actual difference compared with the baseline.
How to get started
For the first meeting, bring one specific task that slows you down, such as a report, a type of meeting, or a kind of request. We assess whether AI Buddy is a fit for it and whether there is something to measure; if not, we say so right away. If you mainly need to teach people to operate a tool, we will recommend training. No commitment and no presentation, just questions about your work.
Pick one piece of work that slows you down
One specific task is enough for the first meeting. No commitment.