Our work

A stock answer should not take a round of phone calls.

The Warehouse Operations Agent shows how stock information and company rules can be brought into one conversation. Here is the business problem behind the build.

Built project / Inventory assistant

Answer stock questions without the back-and-forth.

When a customer needs an answer about stock, your team may have to check the stock system, look up a supplier’s terms and ask a manager about the rules. That is time the customer spends waiting.

GenSage’s Warehouse Operations Agent brings stock information and company rules into one conversation. Staff can check what is available and whether an emergency order qualifies, without piecing the answer together themselves.

A built demo using synthetic demo data, not a client result. Connecting it to your business is a separately scoped project.

See how it could help your team
Example question / sample data

“We’re running low on maize meal. Can I place an emergency order?”

01 Check the stock and reorder level
02 Check whether the emergency-order rule is met
03 Find the supplier terms and approval rules
Your team reviews the answer. The assistant does not place orders or change stock.

The reason for the project

Keep the customer’s question moving.

The everyday problem

Knowing the stock figure does not always answer the customer’s question. A staff member may also need supplier terms and approval rules before they know what to do next.

What the assistant brings together

It checks the product record, works out whether an emergency order qualifies, and finds relevant company policy. The person using it can review the information and decide on the next step.

What a business would measure

Time spent answering a stock enquiry, the number of separate lookups, and how often a colleague has to step in. A pilot would establish whether the improvement is worth the cost; no client savings are claimed here.

Built with synthetic demo data. It does not place orders or change stock. A connection to your own systems is scoped separately.

For your technical team

The project uses Strands Agents and Amazon Bedrock, with DynamoDB stock tools, Python eligibility checks and policy retrieval through a Bedrock Knowledge Base. Local use is through Streamlit; AgentCore Runtime deployment is documented in the repository.

View source and architecture
Discuss a similar problem in your business

Start with the cost of the problem

What is repetitive work costing your business?

Tell us what takes too long, how often it happens and who gets held up. We’ll help you work out whether fixing it is worth the investment.

Discuss Your Project