
Unleashed - How to Thrive as an Independent Professional 643. Scarlett Jiang, COO at Vantage Global AI Shares 3 Live Client AI Use Cases
Show Notes:
Scarlett Jiang from Vantage AI, an AI product and services firm based in London, provides a one-minute overview of Vantage AI, highlighting their focus on data foundations and AI transformation. Vantage AI helps companies consolidate data from various systems into a single source of truth. Scarlett mentions the firm's experience with hospitality franchise clients, such as Burger King, KFC, and McDonald's.
Mock Demo of Chatbot
Scarlett introduces a mock demo of a chatbot designed for hospitality franchise owners. The chatbot can handle real-time queries about sales data, labor costs, and other key metrics. Scarlett explains the process of using the chatbot to query data, including translating natural language questions into SQL, which means users do not need to know SQL.
Custom Dashboard
Scarlett introduces the custom dashboard with data intelligence analyst chatbot functionality, allowing users to query via human natural language and retrieve insights from pre-ingested data warehouses.
Sales Performance
The chatbot can provide summaries of sales performance, labor data, and other operational metrics.
Sales Performance Rank
Scarlett shows how the chatbot can handle more complex queries, e.g.: If I had to focus on 3 stores to improve performance this quarter, which would you recommend and why? (chatbot showcase the capability to synthesize sales, reviews, and trend data into recommended action)
Performance Graph
The chatbot can provide detailed insights into top and bottom performers, including specific metrics like net sales and transaction counts. Scarlett discusses the benefits of using a chatbot for specific questions, rather than pre-built dashboards. The chatbot can also provide reasoning behind its answers, showing the steps it takes to generate insights.
The Process of Building AI Tools
Scarlett explains the process of building AI tools, starting with a diagnostic phase to understand the client's data journey and use cases. After the diagnostic, a strategic roadmap is created to prioritize use cases. A quick prototype is then developed, followed by data foundation transformation. The process can range from a few days to several months, depending on the complexity and scope of the project.
Accounts Payable Month-end Reconciliation Demo
Scarlett demonstrates a workflow automation tool for account payable month-end reconciliation.
Accounts Payable Reconciliation
This demo presents a finance use case built around month-end accounts payable reconciliation - a process every finance team navigates. Supplier invoice data sits across two systems: the AP subledger, which holds granular invoice-level detail, and the general ledger control account, which carries a single summary balance that should match. In practice, the two rarely align - late-posted invoices and manual journal entries that bypass the subledger are the most common culprits. This demo showcases an AI agent that pulls data from both sources, identifies and reconciles the gap automatically; and surfaces discrepancies to human reviewers for sign-off or overwrite - eliminating hours of manual investigation at close.
Converting PDF Purchase Orders into CSV Files
Scarlett demonstrates a tool that converts PDF purchase orders into CSV files.
Snowflake Tables
The tool extracts key information from the PDF, such as contract terms, payment schedules, and expenditures. The tool can transform the extracted data into a chart format for easier analysis.
Reconciliation Report
Payment Breakdown
The tool is designed to automate the process of working with large amounts of unstructured data, reducing manual effort.
Cost and Development Time
Scarlett discusses the cost and development time for AI tools, noting that prototypes can be developed quickly. The bulk of the work involves data cleaning, ingestion, and transformation to ensure data accuracy. The development time can range from a few days to several months, depending on the complexity and scope of the project. The cost varies based on the specific requirements and the level of automation needed.
Demonstration Videos:
Pre-recorded Demo 1: Data Intelligence demo https://www.vantageglobal.ai/insights/demo-pages/data-intelligence-analyst
Pre-recorded Demo 2: Accounts payable month-end reconciliation agent https://www.vantageglobal.ai/insights/demo-pages/ap-month-end-reconciliation-agent
Pre-recorded Demo 3: Parsing unstructured data to structured data https://www.vantageglobal.ai/insights/demo-pages/purchase-order-explorer-agent
Timestamps:
02:24: Demonstration of AI Chatbot for Hospitality Franchise Owners
07:14: Advanced Query Capabilities of the Chatbot
13:06: Process of Building AI Tools at Vantage AI
17:40: Case Study: Account Payable Month-End Reconciliation
28:03: Case Study: PDF to CSV Transformation
34:42: Cost and Development Time for AI Tools
This episode on Umbrex: https://umbrex.com/unleashed/episode-643-scarlett-jiang-coo-at-vantage-global-ai-shares-3-live-client-ai-use-cases/
Unleashed is produced by Umbrex, which has a mission of connecting independent management consultants with one another, creating opportunities for members to meet, build relationships, and share lessons learned. Learn more at www.umbrex.com.
*AI generated timestamps and show notes.
