Tuesday, October 6, 2026
Codeffekt will present its AI mentor at SIDO in Lyon on September 16 and 17, 2026, at booth E318
ANALYZESIMULATE
Codeffekt will present its AI mentor at SIDO in Lyon on September 16 and 17, 2026, at booth E318.
On this occasion, Codeffekt will present its know-how transmission solution, designed to capture the expertise of teams, structure it into a knowledge base and make it available to employees via an assistant available continuously.
The transmission of know-how: a key to business continuity
The solution arose from an observation on the ground: in many organizations, critical skills rely on a limited number of experts and are not formalized anywhere. Procedures document the prescribed procedures, rarely the reality: the trade-offs, the know-how, and the specific cases accumulated over the years. Their departure represents a loss of operational capacity that is difficult to compensate for.
to reconstruct.
From heterogeneous files to a usable corpus
A company's data is rarely homogeneous: office documents, PDFs, procedures, business data, but also audio and video recordings. Codeffekt processes everything using a unified processing chain.
Each source is first converted into a pivot format, regardless of its original format. Audio and video files are automatically transcribed, making previously inaccessible material usable: expert interviews, technical meetings, filmed training sessions, and demonstrations of techniques in workshops. Next comes the extraction: concepts, procedures, equipment, stakeholders, and relationships are identified in the resulting text and then indexed for searchability.
Knowledge graph and wiki: two views of the same corpus
The extracted elements are not simply piled up in a search engine; they are linked within a knowledge graph, which connects a procedure to the relevant equipment, associated typical incidents, and the expert who described it. This structure makes it possible to track dependencies between knowledge areas, identify poorly documented sections, and provide a complete context rather than an isolated excerpt.
The same corpus is published as a navigable wiki, organized by topic and written in the company's terminology. Each page remains linked to its original source, making the information verifiable and attributable, and experts enrich it as they use it: a helpful tip, a specific case, a missing detail. The graph supports the assistant's reasoning; the wiki gives teams a clear and manageable view of their own knowledge.
A durable asset, independent of tools
This knowledge base constitutes the company's assets and belongs to it. It remains accessible from existing tools, such as Teams or Slack, or from the chosen AI model, which can be replaced without the knowledge base needing to be redone.
This is what underpins Codeffekt's neutrality regarding tools: the use of an AI model occurs after the diagnosis, and only when it provides a measurable gain.
Twenty years of business software engineering
For twenty years, Codeffekt has been designing business software, automating low-value-added processes, and focusing on user interface ergonomics. This foundation explains its demanding approach to architectures: robust, secure, and scalable.
The company is co-founded by Fabien Pélisson, PhD in Computer Science, and Mathieu Razafimahazo, UX/UI expert. Based in the Auvergne-Rhône-Alpes region, it is supported by the Auvergne-Rhône-Alpes Region, French Tech, Digital League and Minalogic.
Live demonstration at booth E318
During the two days of the trade fair, the Codeffekt teams will present the solution in operation, applied to similar situations
from those of the visitors, and will exchange on the critical know-how of each organization.
Come and meet Codeffekt at booth E318