Knowledge work drowns in noise.
This is knowledge work, as we know it. And this is how it looks when knowledge, people and AI are organized.
Decisions
- Supplier chosen
- Budget approved
- Pilot starts in March
- Reporting once a month
- Purchase moved to autumn
Files
- Quote.docx
- Budget 2027.xlsx
- Project plan.pdf
- Draft contract.docx
- Onboarding guide.pdf
AI Harness
- Sovereignty
- Context management
- Autonomous agents
- Local models
- Shared rules
Tasks
- Book the review
- Invite the pilot group
- Update the guide
- Check the draft contract
- Send the quote by Friday
Where to start
The product
Aegis
The organisational brain: a platform for complex knowledge work, where people and AI work from the same knowledge.
Discover Aegis
Working with us
Services
Consulting, workshops, talks and projects. The prices are in view.
See the services
The people
Team
Three founders: AI in knowledge work, local AI infrastructure, and security.
Meet the teamAegis
The organisational brain
Aegis is a platform for complex knowledge work. It holds your organisation's shared knowledge and context in one place: the decisions, files, conversations and tasks that people and AI both work from.
It is built to lessen cognitive load and to keep context in order, for humans and AI alike.
Locally or in the cloud
The Aegis demo version: a person's own page with the day's tasks, the group chat and a document, with the AI in the same views.
Pick an answer. The research says the rest.
These three questions come from our quiz. Pick the answer that describes your organisation, and see what the research says about it and which sources it rests on.
When people in your organisation need information, past decisions, or expert knowledge to do their work, how easily can they find it?
What the research says
Mixed findability means a meaningful part of the week is still spent searching or asking around, the pattern that information-overload research links to stress and lower performance (Arnold, Goldschmitt, & Rigotti, 2023; Nakash & Bouhnik, 2024).
The sun rises as the practice gets stronger.An example answer is selected. Pick your own.
Sources
- Arnold, M., Goldschmitt, M., & Rigotti, T. (2023). Dealing with information overload: A comprehensive review. Frontiers in Psychology, 14, Article 1122200. https://doi.org/10.3389/fpsyg.2023.1122200
- Nakash, M., & Bouhnik, D. (2024). How much time does the workforce spend searching for information in the "new normal"? Paper presented at iConference 2024. IDEALS, University of Illinois Urbana-Champaign. https://www.ideals.illinois.edu/items/129980
Sources (2)
- Arnold, M., Goldschmitt, M., & Rigotti, T. (2023). Dealing with information overload: A comprehensive review. Frontiers in Psychology, 14, Article 1122200. https://doi.org/10.3389/fpsyg.2023.1122200
- Nakash, M., & Bouhnik, D. (2024). How much time does the workforce spend searching for information in the "new normal"? Paper presented at iConference 2024. IDEALS, University of Illinois Urbana-Champaign. https://www.ideals.illinois.edu/items/129980
Four ways to work with us
We listen first. Then we design the work together with you, and deliver what was agreed.
By the hour
01
Consulting
Expert work by the hour, each of us in our own field.
On stage
02
Talks
Elina and Petteri speak at conferences and events.
Learn by doing
03
Workshops
Two workshops led by Elina: specification-first AI-assisted coding, and context management in organisations.
Planned together
04
Projects
Larger pieces of work are planned together.
Sovereign AI
Local AI keeps your data in your own network
Under the US CLOUD Act (2018), a US provider can be compelled to disclose data in its possession, wherever in the world the data is stored. A local language model sends no data outside your organisation.
This matters when you process data covered by the GDPR, or data you do not want to hand to a cloud service.
Source 18 U.S.C. § 2713 (CLOUD Act)
A cloud service in the United States
The data leaves
Start
Your document
- prompt and attachments · across the border
Service
Cloud model
Can be compelled to disclose
US authority
Local AI
The data stays
Start
Your document
- inside your own network
Your own server
Local model
Three founders, three fields
Logean was founded in January 2025. There are three of us, each with a field of our own.
AI in knowledge work
Elina Heikkinen
Co-founder, CEO · Product Architect
Knows what today's AI tools can and cannot do, and turns them into better knowledge work.
AI infrastructure
Petteri Heikkinen
Co-founder, CTO · IT Architect
Designs the environments that heavy computing and local AI run on, and keeps them running.
Security
Karoliina Kemppainen
Co-founder, CSO
Makes sure software and AI are built and used securely, from architecture to daily practice.
Aina
Coming
How to AI: AI explained in a few minutes
Aina, the Logean avatar, teaches AI-related concepts in short, easy-to-digest videos.
The series will be published on social media, first on LinkedIn. It has three levels: beginner, intermediate and advanced.
Email address
Free 30-min call
Let's talk for half an hour
A free 30-minute call commits you to nothing. Send a request, and we reply by email to agree a time.
Or ask first
Our chatbot answers questions about Logean. It runs on Logean's own server.