Data and connections
The client’s internal AI assistant and tools for documents and data, with use cases chosen for that environment.
From a niche curiosity to everyday work
An adoption programme after which the company's internal AI assistant became a tool for everyday work: adoption grew from 10% to 87%, and daily active users from 30-40 to more than 160.
Results
- 10% → 87%
- adoption of the internal AI assistant
- 30-40 → 160+
- daily active users
- 30
- AI ambassadors from more than 10 departments
How it works
Ambassadors from 10+ departments
30 people
Consultations and workshops
work in small groups
Catalogue of use cases
approx. 23, built with teams
Only from approved sources
“I don’t know” is allowed
Measured with tool data
and a before-and-after survey
30 AI ambassadors
21 1:1 consultations
7 hands-on workshops
1 Q&A session
The assistant in everyday work
- Emails
- Finding information
- Analysing multiple files
- Working in Excel
What is our department’s budget for next year?
I don’t know: this information is not in the department’s documents.
Result10% → 87%adoption of the internal AI assistant
01
Challenge
Employees had access to an internal AI assistant, but only a small part of the team used it regularly. Company policy restricted public AI tools, and medical data and regulatory requirements called for full control. The company wanted the assistant used regularly and every department to have a list of use cases in its own work.
Access alone did not change everyday habits.
02
Solution
We did not build a new tool. We ran an adoption programme for the one the company already had. We selected and trained 30 ambassadors from more than 10 departments, ran 21 one-to-one consultations, a Q&A session and 7 hands-on workshops in subgroups. Together with the teams we built a catalogue of approx. 23 use cases for the assistant in everyday work, from emails and information search to analysing multiple files and working in Excel.
03
How we built it
The programme ran only in the client's tool and within its security policy: trainers worked on a contractor account, with no access to company systems. Participants built their own knowledge bases in the assistant from department documents, plus ready-made instructions for repetitive tasks, including Excel macros written with AI help. The training covered reducing hallucinations: answers only from specified sources and permission to answer “I don't know”. We measured progress with tool data (daily active users, number of queries) and a survey before and after the programme.
Client's AI assistantKnowledge bases (RAG)Excel and VBAPower AutomatePerplexity
Workflow
How the solution works
- 01
Choose use cases
Ambassadors and employees identify tasks suited to the approved tool.
- 02
Practise and consult
They test work on documents and data, with support to bring use cases into daily routines.
- 03
Check adoption
Usage data and surveys show whether the new ways of working are taking hold.
The human role
Participants verify outputs and ambassadors support practice across departments. The company uses its own approved environment.
Lessons
What made the difference.
- 01
Ambassadors in departments
30 people from more than 10 departments went through workshops, and in 21 one-to-one consultations we mapped their tasks and quick improvements.
- 02
Department knowledge bases
Participants connected department documents to the assistant and got answers based on them. When there is no source, the model is to say “I don't know”.
- 03
Usage, not account numbers
We tracked progress by daily active users and the number of queries in the assistant, not by the number of accounts assigned.
AI adoptionAI ambassadorsUsage measurement
Company of the Future
How this builds the Company of the Future
A project delivers results on its own, but it also leaves organized context behind in the company. Every next AI solution builds on it. This project added to the AI Engine:
- Human+AI Teams
30 AI ambassadors from more than 10 departments.
- Skills
A catalogue of approx. 23 assistant use cases and ready-made instructions for repetitive tasks, written together with the teams.
- Context Lakehouse
Knowledge bases built on department documents, which participants created in the company’s assistant.
Value
Where value is created
Time saved
Employees use the assistant for repetitive document and data tasks.
New possibilities
More departments can identify and test AI use cases independently.
What comes next
The outcome is broader, regular assistant use and departments better able to assess further applications.
Most of our projects are covered by NDAs, so we only share part of them, without client names or details, as inspiration. We are happy to tell you more about our experience in a conversation.
Let’s talk about the work you want to improve.
Tell us how a process works today. Together, we will explore where AI Engine can help and which module to start with.
Book a callOr write to: hello@wearefuture.ai
