How we improved operational management with generative AI and a cloud-native architecture on AWS
The challenge
In retail, day-to-day operations never slow down: shift changes, unexpected absences, peak demand periods, and teams spread across different departments. In that context, being able to quickly know who is available is not a “nice to have” — it is an operational necessity.
Our client, a leading retail company with stores across multiple locations, faced a very specific challenge: store managers needed a fast and reliable way to check staff availability, by store and by department.
Up until then, the process was slower than it should have been and created friction in daily operations:
-
Manual and repetitive queries: managers had to check multiple systems to find schedules, shifts, and availability.
-
Wasted time: managers spent too much time searching for information instead of focusing on operations and customer service.
-
Slow responses: there was no quick way to retrieve real-time information.
-
Scattered information: data was spread across different tools and formats, making access even more difficult.
In short: the information existed, but finding it was the real challenge.
The solution: a genuinely useful conversational bot
To solve this, we developed an intelligent virtual assistant, integrated directly into the corporate communication tool the client was already using day to day.
This was key: instead of forcing teams to log into yet another platform, the bot lives in the channel where they already work. This enables natural adoption and keeps the learning curve minimal.
The bot allows managers to ask questions in natural language such as:
-
“Who is available tomorrow afternoon in the electronics department at the Madrid store?”
-
“Which employees can cover the night shift this week at the tills?”
The result is simple: managers ask as if they were speaking to a colleague… and receive a clear answer within seconds.
Technical architecture: AWS as the foundation for scalability and performance
The solution was built on a cloud-native architecture designed to be scalable, resilient, and ready to grow alongside the business.
These were the main AWS services used:
OpenSearch retrieves the most relevant information and Bedrock generates the response using that data as its foundation, ensuring the bot answers accurately using real information, not “made-up” content.
How does it work in practice?
From the user’s point of view, the workflow is straightforward — but powerful in execution:
-
The manager submits the query via the corporate chat.
-
Bedrock analyses the intent and identifies key elements such as store, department, date, and shift.
-
OpenSearch searches for the most relevant information using semantic search.
-
Bedrock builds a clear response, contextualised and ready to use.
-
The bot responds immediately in the same chat, without the user needing to switch tools.
Results achieved
The impact was fast and measurable:
→ 80% reduction in the time required to check staff availability.
→ Immediate access to up-to-date information, without leaving the team’s usual working channel.
→ High internal adoption, thanks to direct integration into the corporate tool.
→ Automatic scalability, ideal for handling peak demand during high season.
→ Improved operational decision-making, by making critical information available within seconds.
Conclusion
This project demonstrates how the combination of generative AI, semantic search, and a cloud-native architecture can transform an operational process that was previously slow and fragmented into an agile, simple, and efficient experience.
The success of the project was built on three key pillars:
-
User experience: the bot is available where teams already work, with no friction.
-
Modern managed technology: AWS enabled faster development without adding operational complexity.
-
Focus on real impact: this was not about “implementing AI”, but about solving a real business problem with measurable results.
Would you like to apply generative AI to operational processes within your organisation?
Let’s talk. We can help you identify real use cases and turn them into scalable solutions with direct business impact.