The situation
Employment questions depend on the situation behind them.
Loading the pageLoading the pageLoading the pageEmployment-law intelligence, connected to the matter
ACW connects Leo’s matter-based dialogue, knowledge retrieval and source-linked responses with Azure application delivery. The product helps businesses work through employment matters with a person responsible for reviewing the guidance and deciding what happens next.
Leo is an employment-law intelligence product for businesses, HR leaders and HR consultants. It brings the facts of a matter together with relevant legislation, case material and the organisation’s own documents. Its current proposition extends beyond a general HR chatbot into a workflow for preparing, documenting and reviewing employment decisions. The product supports structured matter intake, source-backed guidance and records that can be taken forward for internal review or to a qualified lawyer. ACW’s confirmed contribution centres on application engineering, retrieval, dialogue and the Azure delivery foundations.
Employment questions depend on the situation behind them.
An answer depends on the source that applies to it.
Guidance depends on the information the user provides.
Leo helps teams work through employment matters with the relevant facts, documents and sources in view. ACW’s application work connects matter-based dialogue with knowledge retrieval and source-linked answers, supporting preparation that a person can check before acting. Product planning, security review and workflow measurement accompanied the AI-product delivery. The result: source-linked answer preparation connected to the employment matter.
Describe the employment matter and provide its context.
Retrieve relevant material and prepare a source-linked response.
Check the response and its references before deciding the next step.
ACW completed the Azure infrastructure and frontend/backend CI/CD foundations for Leo. The release workflow supports the employment-matter application as its retrieval, response and surrounding product behaviour continue through validation.
Prepare the frontend and backend deployment environments.
Connect the code delivery pipelines.
Deploy through the established Azure workflow.
Azure infrastructure and CI/CD delivered for the application.
Validation covers source references as well as upload, API, retrieval and workspace paths around the assistant. Citation correctness is checked separately from the surrounding workflow, while responses remain subject to human review. The result: source-citation checks paired with workflow validation.
Answer preparation and source checking took 10 minutes instead of 25.
The same 30 anonymised scenarios were assessed with specialists.
Specialists confirmed correct source citations in 47 of 50 answers.
