Skip to Content

Has your team already solved that problem? The invisible tax of relearning ends here

Discover how to keep knowledge from getting lost so your team stops spending hours solving what they already knew, using a semantic graph that connects, not just accumulates.
May 30, 2026 by
Has your team already solved that problem? The invisible tax of relearning ends here
Wilfredo Fernando Pastor Avila
| No comments yet

The story of the 3 lost days (and how not to repeat it)

A few months ago, one of our senior consultants spent three full days researching and testing the best way to connect Odoo with Power BI. It was not wasted time: he found a solid, tested and efficient solution. But the real problem started when, a week later, another consultant opened a similar ticket and, without knowing it, started from scratch. He spent another three days on the same research. And then another. And another. In total, that quarter, five people on my team spent 15 person-days rediscovering what had already been discovered. 60 days a year. Three months of one person's work, thrown in the trash. This is the invisible tax of relearning, and most companies don't even measure it.

The solution: a living semantic graph, not a dead folder

What I learned from that experience is that knowledge should not live in loose files, personal notes or static wikis. It needs a home that connects it by meaning, not by file name. A semantic graph that understands the relationships between concepts, even if nobody wrote them explicitly. At Ganemo, we use GitHub as the source of truth and a tool called Blenow that turns that repository into a shared brain for the whole team. It is not a simple folder that accumulates documents; it is a graph that composes and connects every new learning with what already exists.

Real example: When our consultant finished researching the Odoo-Power BI integration, he didn't have to write a manual from scratch. He simply told his AI agent: "Crystallize everything learned about Odoo and Power BI into the organization's brain". The agent, connected to Blenow, first searched the graph to see whether similar knowledge existed, found that it didn't, and proceeded to write the documentation directly in the GitHub repository, with structure, detailed steps and version control. All in under 5 minutes. No manual intervention, no risk of duplication or misplacement. With Ganemo, this workflow is available to any company that wants to implement it.

Step 1: The source of truth on GitHub

It all begins with a GitHub repository that acts as the team's single source of truth. Files are stored there in Markdown format, organized by topic. The advantage: it is versioned, auditable and portable. The knowledge belongs to your company, not to an external tool.

Real example: In our case, we created a specific folder for "Odoo-Power BI Integration". We didn't mix it with the general Odoo documentation because we need granularity in access permissions. The folder is initially empty, but ready to receive the knowledge that gets generated. Team members know that any relevant discovery must end up there, but they no longer have to do it manually: the agent does it for them.

Step 2: Blenow turns GitHub into a semantic graph

Blenow is the service that connects to your GitHub repository and turns all the content into semantic vectors. This means it searches by meaning, not by keywords. When an agent needs to document something new, it first queries Blenow to know whether related information already exists and in which folder it should go. This avoids duplicates and ensures new knowledge integrates harmoniously with existing knowledge.

Real example: When our consultant asked to crystallize the knowledge, his agent used the Blenow MCP. First it searched the graph: "Searching for repositories with similar meaning... Found that the Odoo-Power BI integration repository exists and is empty. No duplicates in other repositories. Proceeding to write." Then the agent generated the complete documentation, including detailed steps, scripts and a table of contents, and uploaded it automatically to the right folder. With Ganemo, setting up Blenow takes minutes and the team starts benefiting immediately.

Step 3: The agent documents for you

Once the agent has access to the graph, any team member can say: "Crystallize what we learned today about [topic]". The agent takes care of everything: searching existing information, writing, structuring, and uploading to the repository. It doesn't matter whether you use Cursor, Cloud Code or Codex; as long as the agent has the Blenow MCP configured, it works the same way.

Real example: After an intensive debugging session, a developer tells his agent: "Save the solution to the Odoo-Power BI connection error in the organization's brain." The agent searches the graph, finds that a document about common errors already exists, and updates that section with the new finding. No manual intervention, and no risk of the information being left in a personal note. Also, since the search is semantic, the agent knows exactly where to place the information, even if the topic name doesn't match exactly.

Results: The end of the invisible tax

With this system, the team's knowledge is no longer lost. When a new member needs to know how to connect Odoo with Power BI, they just ask their agent and get the answer based on the crystallized source of truth. They don't need to research from scratch, ask colleagues, or dig through disorganized folders. Moreover, token savings are significant: by searching with semantic vectors, the agent doesn't have to read entire documents, only retrieves the relevant parts.

Real example: In our company, after implementing this workflow, we measured that over one quarter more than 15 days of duplicated research were avoided. Multiplied over the year, that's 60 person-days saved. This doesn't count the time saved by not having to document manually. With Ganemo, we help companies set up this system in an afternoon, and the results are immediate.

How to start today

Implementing this system requires three pieces: a GitHub repository (your source of truth), the Blenow service (which turns it into a semantic graph), and an AI agent with MCP configured. At Ganemo, we guide you step by step so your team stops paying the invisible tax of relearning. The knowledge you already gained shouldn't have to be earned again.

Related video

Has your team already solved that problem? The invisible tax of relearning ends here
Wilfredo Fernando Pastor Avila May 30, 2026
Share this post
Tags
Archive
Sign in to leave a comment
Does your company create projects manually? Discover how Odoo automates management with templates
Learn how to create projects in Odoo automatically from sales or templates to save time and avoid errors.