Build a Tool Calling AI Agent for Warehouse Stock Transfers
Group: Business Requirement
|Product Category: AI & Data Science
|Sub Category: AI Agent Engineer
Solution Available
Source code, datasets, and other related files for the solution will be provided.
About this Product
Ironvale Manufacturing Co. operates two warehouses that continuously move raw materials and finished components between production lines. Today, coordinators submit stock transfer requests by phone or spreadsheet, while a supervisor manually checks inventory before approving unusual requests. The process is slow, inconsistent, and provides no reliable record of who validated or approved each transfer.
In this lab, you build StockPilot, a single-workflow AI agent that converts a plain-language request, such as “move 40 units of SKU-1001 from WH-A to WH-B,” into a structured, validated action. The agent uses an LLM tool-calling interface with a fixed set of tools for checking stock, validating business rules such as valid SKUs, distinct warehouses, and sufficient inventory, gating high-impact transfers behind human approval, executing approved transfers against local JSON inventory files, and recording every request in an append-only movement ledger.
An explicit workflow-state object tracks each request through its required steps in the correct order. A swappable LLM client allows the workflow to run deterministically with a scripted stub for automated testing, requiring no live API key or network access.
You will practice core AI-agent concepts including the tool-calling loop, fixed tool sets with JSON schemas, explicit workflow state machines, enforced step ordering, human-in-the-loop approval gates, graceful tool-call error handling, business-rule validation, atomic inventory updates, and structured logging for observability.
By the end, you will have a working command-line agent with a passing automated test suite, sample run logs, and a short skill-mapping write-up connecting each AI-agent concept to the code that demonstrates it. These deliverables can be explained in interviews and showcased in your portfolio.
Resources
The complete Warehouse Stock Transfer Business Requirements Document (BRD) is available as a Read-Only attachment. It provides the complete project specification, including the business context, objectives, scope, requirements, use cases, domain concepts, and design expectations. Review the BRD carefully before starting the implementation.
Asset Contents:
Project overview and business objectives
In-scope and out-of-scope definitions
Functional and non-functional requirements
Key use cases and business workflows
Domain model, entities, and core concepts
Design and implementation requirements
Requirements traceability and supporting documentation
Project Mentors
Similar Products
Product Performance Dataset
Topics: SQL, PostgreSQL, Retail Performance
Basic Professional Data Analysis
Topics: SQL, PostgreSQL, Data Quality Analysis
Restaurant Performance & Menu Optimization
Topics: SQL, PostgreSQL, Data Analytics
Similar Services
Finding the best experts for you...
No Services Yet
Expert services for this product will appear here once available.
Top User Reviews
Loading reviews...
Be the first to review this product!
Please try refreshing the page.