Skello uses Amazon Bedrock to query data in a multi-tenant environment (logical boundaries)
Skello, a France-based HR workforce management SaaS platform, built an AI-powered assistant for end users to query workforce data in natural language while preserving multi-tenant data isolation and GDPR compliance. The assistant uses Amazon Bedrock, Amazon Bedrock Guardrails, and AWS Lambda to translate questions into structured database queries, enforce role-based access controls, and generate visualizations automatically.
- Organization
- Skello
- Industry
- Tech & Comms
- Location
- France
- Published
- September 2025
Reported outcomes
Strategic outcomes
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Skello
- Provider
- AWS
- Maturity
- Unknown
- Linked source
- AWS Machine Learning Blog
No explicit deployment-stage evidence found.
Primary read
Use case focus
Showing 3 of 3
- 1Conversational assistants
- 2Decision support
- 3Operational analytics
- Users needed natural-language access to complex workforce data across many tenant customers with correct intent understanding and secure, GDPR-compliant enforcement of data access boundaries.
- Users also needed graph and visualization outputs without writing SQL or using BI tools.
- Skello built an AI assistant that converts natural-language questions into structured database queries.
- The system validates and enforces security policies and role-based access before LLM use, keeps user context and permissions outside the LLM, and uses Guardrails to reduce prompt injection and inappropriate content.
- The assistant executes only authorized queries and automatically generates charts and graphs.
- The solution provides faster, more accurate answers to complex workforce questions.
- It improves usability by avoiding SQL and BI tooling.
- It generates professional-quality charts automatically.
- It maintains strict tenant isolation to prevent cross-contamination of customer data.
Architecture
A serverless multi-tenant assistant architecture uses authentication and authorization services before any LLM call, passes only approved context to Amazon Bedrock, validates queries through a security layer, applies Amazon Bedrock Guardrails, and executes database access only within the user’s permitted scope. The system also generates visualizations automatically from query results.
Sources & evidence1
- Customer explicitly identified
- Technical implementation details available
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?
Community
Comments
Loading comments...