[ MANUFACTURING_/_INDUSTRIAL ]
Automation-Backed Internal Assistant
What it is
An internal chat assistant wired to the company's existing automation platform.

[ THE_PROBLEM ]
Why this existed
Companies that have already built automation get told the AI layer means moving that logic into an application only developers can change.
[ WHAT_WE_BUILT ]
What we built
A chat interface and analytics dashboard backed by n8n workflows rather than direct model calls — so the business logic stays in the automation platform the operations team already owns and can change. Authentication, persisted chat storage and a debug surface, with the API routes, auth and storage layers unit-tested.
- Chat interface backed by existing automation workflows rather than direct model calls
- Analytics dashboard
- Authentication and persisted chat storage
- Debug surface for inspecting behaviour
- Unit-tested API routes, auth and storage layers
[ HOW_IT_IS_USED ]
How a company uses it
A company that has already invested in automation gets a conversational front door to it, without moving the logic into an application only developers can edit.
Built with
[ COMMON_QUESTIONS ]
Questions clients ask
Why route through an automation platform instead of calling the model directly?
Because it keeps the business logic where the operations team can change it. Moving that logic into application code means every future change becomes a development ticket — which is a step backwards for a team that deliberately invested in automation they own.
Does the extra hop hurt performance?
It adds latency, and for a chat interface that is usually acceptable. The trade is speed against maintainability by non-developers, and for internal tools maintainability normally wins.
Why unit-test an internal tool?
Because internal tools have long lives and infrequent attention. Tests on auth, storage and API routes are what let someone change it confidently a year later.
[ RELATED_WORK ]
Similar builds
Enterprise analytics
AI Business Intelligence Workspace
Upload data, get dashboards, ask questions of them in plain language.
Read case studyInvestment research
AI Research Report & Deck Generator
Source documents in; a structured research report or presentation out.
Read case studyInvestment research
Drag-and-Drop Report Builder
A visual editor where analysts compose print-ready reports themselves.
Read case studyIs this close to your problem?
Most engagements start with a version of something on this page. Tell us what is different about yours and we will tell you what it changes.
Start a conversation