People first, then the work, then the tool.
Most companies get AI adoption the wrong way round. They hire an engineer or buy a tool, and leave out the largest group in the building: the people who do the work. When those people understand AI, they can say which process is worth automating, and they own what gets built. Nobody has to guess. That is why we usually start by teaching.
Building first is faster to sell and worse to own. It puts a system in front of a team that cannot say why it exists, and will not defend it when it breaks. So every build we do starts by looking at the work first.
One promise: we tell you honestly whether it is worth it. Even when you have already decided to build.
1 Teach
Your team learns to work with AI.
What AI is, how it thinks, how to brief it, how to check what it gives back, and when not to trust it. Hands-on, and linked to their own work.
There are five ways to do it, from a four-hour session for a whole room to a two-day programme with real work in between.
The AI programme, start to finish.
BEFORE
We agree with you what the team will work on.
DAY ONE
How AI works and how it thinks. Hands-on, in pairs, linked to their own tasks.
WITHIN 48 HOURS
An action plan for the team.
FOUR WEEKS
Your team tries it on real work and collects its questions.
DAY TWO
Starts from what they tried, and goes deeper on their questions.
AFTER
A record you keep of what was covered.
The other workshops are shorter versions of the same idea.
2 Assess
Together, we find where it is worth it.
We sit beside the people doing the work and map it as it really runs: which processes exist, who touches them, where the hours go, and which steps need human judgement. We also collect the ideas your people already have, and the new ones that come up once they start seeing what AI could do for their work.
Every opportunity is scored on impact and effort, so you can see why one comes before another, and argue with it. Every recommendation shows its return: the hours it gives back in a year, and what those hours cost you today, worked out from your own numbers. When you ask us to build one, the written quote sits next to that figure, so you see when it pays for itself before you sign anything.
The report is written in plain prose, not slides, so you can take it to whoever holds the budget. Every figure shows how it was worked out, so your finance team can check it.
What you get
- Time on site, beside the people who run the processes
- A written assessment report, with the timing agreed up front
- The return on every recommendation, in hours and euros a year
- An answer to every idea your team raised
- A Not Recommended section, always: what is not worth it, and what is not yet
See what an assessment report looks like.
A complete sample for a fictional company: how the work was mapped, the opportunity matrix, the recommendations with what each one gives back, and the Not Recommended page.
Email hello@tegron.eu and we send you the sample report.
Cover
Opportunity matrix
Not RecommendedWe will tell you what not to automate.
Every no comes with a reason. Some are final. Some are not yet, and then we tell you what would have to change first. That includes the ideas your own people raised.
TWO EXAMPLES FROM THE SAMPLE REPORT
NOT WORTH IT
AI replies to complaints
Few complaints, high stakes. Every one deserves a person. Keep it human.
NOT YET
Automatic lead scoring
Half the CRM records are incomplete, so any score would be a guess.
WHAT WOULD HAVE TO CHANGE
Three months of complete CRM records. Then we look again.
3 Build
Together, we build what is worth building.
Every build starts by looking at the work, with your team involved from the first specification to the handover. The specification comes first, so there is something to disagree with before any code exists.
A working prototype comes early. You sign it off in writing, or we change the design. It runs on your systems, within the rules your own IT sets.
What you get
- A specification you sign off before any code is written
- A working prototype early on, with written sign-off
- The production build, with documentation and a handover session
- Your team able to run it and change it without us
- 30 days of support after handover
Rapid Deployment, on request
Your engineers build. Your teams know the work. We translate.
For companies with their own engineers, we work team by team. For the first two days we sit beside the work and teach the team what AI can do for it. Then, together with that team, we build a working prototype in short, intense sessions, within the rules your engineers set. The prototype shows your engineers exactly what is needed and what it is worth, before they spend a day on it. They build it out properly, and we stay in the middle, translating between the business and the build. Then the next team. We discuss it on the call.
YOUR TEAM
knows the work
TEGRON
teaches and translates
YOUR ENGINEERS
set the rules and build
Yours, not ours.
You own what we build
The code, prompts, workflows and documentation we build for you are yours on final payment. It is in the contract before the work starts.
Your data stays yours
Everything you share is confidential and used only for your work. Never reused, never shared. We keep it only as long as the work needs it, plus what the law requires for our own records. A processing agreement is signed before any project that touches personal data.
Your team can run it
Every build comes with documentation and a handover, so your own people can change it without calling us.
Your data, in detail.
- Everything you share stays confidential, in both directions.
- We never share or reuse your data, your documents or your business information. We record a session only if you agree.
- On a training day, your people work in AI accounts your company holds, on their own laptops.
- A build runs on your systems, with credentials you provide. We host nothing unless you ask.
- Documents you give us are processed only in AI tools we name in writing before we start, and deleted once the work no longer needs them, or sooner if you ask. A processing agreement is signed before any project that touches personal data.
Tell us how your work runs.
Thirty minutes, free, and not a pitch, with one of the two founders. If AI is not the answer, you will hear that too.