How much time does your team spend searching for information within the ERP, updating records, reviewing opportunities, consulting documents, or preparing reports that could already be available in the system?
In many companies, the problem is no longer the lack of information. The problem is all the work we still do to find it, interpret it, and turn it into action. This is where Odoo 20 and MCP start to change the way we understand an ERP.
With the evolution of artificial intelligence in Odoo, AI agents are no longer designed just to answer questions or generate content. They can also use system tools, consult information, work with documents, and, when the process allows, create or update records.
Odoo summarizes it in a fairly simple way: its AI agents are evolving from answering questions to doing work.
And for a company, that difference can be much more important than simply incorporating another chatbot.
Odoo 20 and MCP: what really changes for a company?
Imagine that a sales director needs to know which opportunities should close this month. Today he would probably have to enter the CRM, apply filters, review sellers, compare dates, and analyze which opportunities have been without follow-up for too long.
With a properly configured agent, the interaction can come much closer to this:
“Show me the opportunities that should close this month, identify which ones have been inactive for more than seven days, and tell me which ones need attention.”
Artificial intelligence can interpret the request, consult the information available in Odoo, and present the result in a much simpler way for the user. But the real change comes when AI can also use tools.
In Odoo 20, agents can work with capabilities to consult information and, depending on the configuration and permissions set, perform actions such as creating records, updating information, or working with documents.
This opens interesting scenarios in practically all areas of a company:
- Sales: detect follow-up opportunities, consult information about customers, or assist in creating and updating records.
- Customer service: analyze tickets, identify recurring issues, and consult internal documentation to respond faster.
- Administration: locate information without navigating through multiple menus, filters, and reports.
- Documents: consult the content of files and relate it to information existing within the ERP.
- Management: ask questions about operational information using natural language and get a response based on the data available in Odoo.
- Operations: support processes where currently a person must review information before deciding what action to take.
This means that with the help of Odoo, people can spend less time telling the system where to look and more time deciding what to do with the information they find.
With AI agents, an additional layer appears: now the user can express an intention in natural language and allow the agent to determine what information or tool it needs to use.
And what does MCP have to do with all this?
Here comes the Model Context Protocol, better known as MCP.
Although the name may sound technical, the idea behind it is much simpler.
MCP allows applications and artificial intelligence agents to connect with external systems through a standardized way of discovering and using the information and tools that those systems make available.
In the case of Odoo 20 and MCP, an Odoo database can function as an MCP server to which an external artificial intelligence agent connects.
In other words, MCP works as a bridge. On one side we have the information and processes of the company within Odoo. On the other, we have artificial intelligence agents capable of understanding instructions in natural language.
MCP helps both worlds communicate.
A user could request:
“Analyze the open opportunities from the last three months and tell me which have the highest amount but least commercial activity.”
The agent can use the tools that Odoo has allowed them to consult, obtain the necessary information, and use it to build a response.
In another scenario, they could request:
“Create an opportunity for this new client and assign it to the corresponding sales team.”
If there is an authorized tool to perform that operation and the user has the necessary permissions, the agent can use it.
But there is an important point: MCP does not mean handing over your entire database to an artificial intelligence without control. Connections are authenticated and the available tools can be managed. Additionally, Odoo maintains its own permissions and access rules.
This allows for the creation of much more controlled scenarios.
An agent designed for the sales team could consult opportunities and clients without needing access to sensitive information from other areas. Another agent specialized in support could work with tickets and internal documentation without having permissions over the accounting.
This is where the implementation becomes really important.
Because connecting AI with an ERP is relatively easy to explain. The difficult part is determining what it should be able to do, what information it can consult, and which processes are worth automating.
Having artificial intelligence is not the same as leveraging it.
A company does not gain a competitive advantage simply by activating an artificial intelligence function. It gains it when it correctly identifies where AI can eliminate friction from its processes.
For example, automating a task that a person performs for two minutes once a month will probably have little impact.
But using AI to help a sales team detect forgotten opportunities daily, summarize conversations, and find relevant information can have a much greater effect. The same goes for documents, customer service, administration, or operational analysis.
With Odoo 20, agents and MCP take that idea even further. It’s no longer just about generating content or answering questions. It’s about starting to build a new way to interact with the ERP.
One where the user can express what they need in natural language and Odoo can help them turn that intention into information, analysis, or concrete actions.
Turns Odoo into a tool that truly works for your business
The arrival of artificial intelligence agents and MCP opens important possibilities for companies that want to reduce manual tasks, better leverage their data, and simplify the way their teams work with the ERP.
But the real value of Odoo 20 and MCP is not in connecting an artificial intelligence to your database.
It’s in correctly designing which processes need improvement.
In Business Insights we are official partners of Odoo and we help companies implement, customize, and automate their ERP according to their actual operational processes.
We analyze how your company currently works, identify tasks that can be simplified, and design solutions using Odoo, automation, and artificial intelligence when they truly add value.
Because implementing technology should theoretically not mean adding more complexity.
It should mean selling better, operating with greater control, reducing administrative work, and making better decisions with the information that your company already generates.
Do you want to discover which processes your company could improve with Odoo 20, automation, and artificial intelligence?
Contact us and schedule a consultation with Business Insights. Let's analyze together how to turn your ERP into a platform that not only stores your company's operations but also helps make it more efficient.