By: Mauro Soto - November 14th, 2026; Technology Manager, Animistic Â
On October 22, I had the opportunity to attend the AI, Context and Business Impact conference in San Pedro Sula, Honduras, presented by Strategy and BACG (Business Analytics Consulting Group).
The conference introduced Strategy Mosaic, a solution designed to connect artificial intelligence with business data. However, what caught my attention the most was not the technology itself, but something that I believe is extremely important when implementing AI: context.
Artificial intelligence has become a powerful tool for businesses. However, one of the challenges today is that we can ask the same question multiple times and receive different answers.
For example, imagine asking an AI agent for the total sales of a company. Does that include taxes? Discounts? Returns? Or is it simply the total amount before any adjustments? The information may already exist in a database, but that does not mean the AI understands how the business defines or calculates it.
From my perspective as a software engineer, I believe the problem is not necessarily AI itself, but how we provide information and instructions. We often know what we want to ask, but we do not always explain the context behind our questions. AI generates responses based on the information and instructions it receives. Without properly defined context, it may produce incomplete, inconsistent, or incorrect results.
During the conference, Strategy introduced Mosaic, a solution that connects different databases, applications, dashboards, and AI agents through a shared set of business definitions. Instead of having each AI agent interpret the same information differently, businesses can define how their data should be calculated and used beforehand.
The platform also supports reports, dashboards, AI agents, security controls, and integrations with other AI technologies through MCP (Model Context Protocol, an open standard originally developed by Anthropic that connects AI applications to external data sources and tools).
Another interesting aspect was the pricing model presented. According to the presenters, customers can create and consult unlimited AI agents without separate charges for each input or output token. However, the most important part for me was how Mosaic approaches the problem of inconsistent business information.
At Animistic, we work with companies in animal nutrition, product development, formulation, research, and regulatory support. These are areas where having the correct information and understanding its context is extremely important.
Consider a company developing a new animal nutrition product. It may have formulation information in one system, ingredient costs in another, research results in spreadsheets, and customer information in a CRM. If an AI agent is asked to calculate the cost of a product or analyze research results, it needs more than access to those databases. It needs to understand which formulation is being used, what ingredient prices apply, what species the product is intended for, and which calculations have already been validated.
The same applies to research. An AI agent analyzing animal performance should understand the trial conditions, the measurements being used, and what the results actually represent. Without that context, it could generate an answer that looks correct but does not accurately represent the information. For our clients, this means AI could become a useful tool for analyzing research, reviewing product information, preparing reports, and supporting business decisions, provided the information and business rules are properly defined.
It could also help connect information that is currently distributed across different systems, making it easier for teams to access consistent information without having to review multiple spreadsheets or applications.
However, the goal is not to replace the expertise of nutritionists, researchers, or other specialists, but to make their work more efficient and allow them to focus on what they do best. For example, AI could help a nutritionist review ingredient information, compare formulations, or organize technical specifications, but the nutritionist would still determine whether those formulations are appropriate for the animal and its specific needs. Similarly, researchers could use AI to organize results, identify patterns, and prepare initial reports, allowing them to spend more time interpreting the findings and evaluating their scientific relevance.
In this way, AI becomes a tool that supports our specialists rather than replacing their technical judgment. It can help reduce the time spent searching for information, organizing documents, or performing repetitive tasks, giving professionals more time to focus on research, problem solving, and decisions that require their knowledge and experience.
Over the past year, we have been working on improving our own internal operations at Animistic by building systems that help streamline tasks and connect information across different platforms. This includes processes related to purchasing, receiving, inventory management, and production, where information previously required multiple manual steps across spreadsheets, emails, and applications.
One example is our ongoing work to connect purchasing information with receiving and inventory records. A system needs to understand the difference between products that have been ordered, products that are expected to arrive, and quantities that have actually been received. Those differences are important because an item should not be considered available in inventory simply because a purchase order exists. The system also needs to identify discrepancies and allow the appropriate team members to review them before making changes to the inventory records.
Building these systems has involved trial and error, testing different approaches, identifying issues, and making adjustments as we better understand how each process works. What may initially seem like a simple automation often requires defining business rules, understanding how different systems communicate, and making sure the information remains accurate throughout the process.
Not every part of this work involves AI, but the experience has reinforced the same principle discussed during the conference. Technology can help us reduce repetitive work and improve how information moves through a business, but it still depends on having the right context and rules in place.
For us, the objective is not to remove people from these processes. It is to make their work easier, reduce unnecessary manual tasks, and provide more reliable information so they can focus on the areas where their experience and judgment are most valuable.
One thing this conference reinforced for me is that implementing AI is not simply about creating agents or connecting databases. Before doing that, we need to understand the information, define the business rules, verify the calculations, and determine what the AI should be allowed to access. That is where software engineers, data engineers, and industry specialists play an important role.
At Animistic, we already understand how important science, accurate information, and technical expertise are for our clients. Through our own experience developing and improving internal systems, we have also learned that technology becomes more useful when it is designed around how a business actually operates.
As AI continues to become part of business operations, applying that same attention to how information is organized and used will be equally important.
My biggest takeaway is simple: AI needs more than data. It needs context. When that context is properly defined, AI can become a more reliable tool that helps professionals work more efficiently, make better informed decisions, and focus on the expertise that brings value to their businesses.
Â
