At CIM we have been working for years with complex document management environments, especially in platforms such as OpenText Documentum, where information is not only abundant, but also important for the business.
However, there is a reality that we see repeated in many organizations:
Having the information does not mean being able to exploit it in an agile way. Therefore, it is time to apply document exploitation with AI.
Today, accessing a document is relatively simple. But understanding it, cross-referencing it with others, extracting value or answering complex questions is still a slow, manual process, highly dependent on expert knowledge.
That is why we have launched an AI pilot with a clear objective:
Move towards a model in which it is possible to "talk to the documents", regardless of where they are stored.
The starting point
Large organizations have robust document repositories: Documentum, SharePoint, file systems, etc.
The problem is not usually in the technology.
The problem is in the operating model:
- Keyword-based searches
- Manual navigation by folders or metadata
- Dependence on expert users
- Difficulty connecting information between documents
This greatly limits the real value of the information available.
The proposal: an intelligent operating layer
The pilot we are working on introduces an additional layer on top of the existing repositories that allows interacting with the information in a more natural way.
Instead of searching for documents, the user can:
- Formulate complex questions
- Obtain contextualized answers
- Browse knowledge, not folders
What’s more, it does so from an environment that is already part of users’ daily lives:
Interaction is done directly through Microsoft Teams.
This reduces adoption friction by not requiring new tools or relevant changes in work habits.
Repository Independence: Documentum… or any other repository
Although the pilot was initially developed on OpenText Documentum, the approach is repository agnostic:
- Documentum
- SharePoint
- File systems
- Other document repositories
The architecture allows connecting to different sources without changing the user experience.
This is especially relevant in corporate environments where multiple systems coexist.
Security in document exploitation with AI
One of the critical aspects of any AI initiative is security.
From the beginning of the pilot we had a clear premise:
AI must adapt to the existing security model, not the other way around.
Therefore, the approach we are developing:
- Respect the permissions of the source repository
- Verify access in real time
- Filter information based on the user
In environments such as Documentum:
- Existing ACLs are maintained
- The user only accesses the information for which he/she already has permission.
This allows the use of AI to be explored without the need to redesign security models.
Cloud infrastructure: the basis for scaling
The pilot is based on a flexible cloud architecture:
- Distributed Infrastructure on Azure and AWS
- Scalability on demand
- Processing capacity for large volumes of documents
Although we are in the development phase, this basis allows us to validate the approach in real scenarios and with relevant data.
The bottom line change: moving from seeking to understanding
Beyond technology, what we are exploring is a paradigm shift:
Before:
- Search documents
- Read them manually
- Interpreting information
Now:
- Ask questions directly (e.g. from Teams)
- Obtain contextualized answers
- Navigating structured knowledge
We went from managing documents to starting to exploit information.
An evolving approach
This pilot is part of a line of work that we continue to develop and refine.
The objective is not to replace existing systems, but to expand their operating capacity without adding complexity or compromising security.
As with any initiative of this type, there are still challenges to be resolved, but early results point to a clear improvement in the way users access and use information.
Conclusion
Document management alone is no longer sufficient in environments where information really matters.
The value lies in the ability to exploit it intelligently, safely and transversally.
And if this also occurs within already adopted tools such as Microsoft Teams, the impact on the organization can be significant.
If you would like to learn more or explore how this approach could be applied in your organization, we would be happy to share what we are building. If you have any questions, you can consult our website here.
