Inverk.
Back to Inverk / Services
SERVICES

AI assistants

An assistant can help visitors with questions and relevant next steps. We define its knowledge sources and any tool connections together. The model and provider are selected for the use case; secrets stay server-side.

Discuss your project

Delivered functions & examples

  • Website and Telegram assistants
  • FAQ and knowledge assistants
  • Lead qualification and enquiry clarification
  • Documents and business knowledge as agreed sources
  • Optional workflow and tool connections

What we need from you

  • Approved business information, FAQ and the sources the assistant may use.
  • Questions it should answer, limits and when to refer a visitor to a person.
  • Your services, intended customers and the task to solve.
  • Existing content, brand materials and access to relevant systems.
  • A person responsible for feedback and approval.

How we build the project

  1. Clarify the need: Your services, customers, existing systems and the flow you want to improve.
  2. Agree the scope: Deliverables, price, milestones and review points in a written offer. A prototype can help clarify the direction where appropriate.
  3. Build and review: Review working previews against the agreed scope and resolve feedback at each milestone.
  4. Launch and hand over: Agree publication, access, source delivery and setup documentation relevant to the project.

Handover

  • Inverk-managed deployment can be agreed, or the solution can be deployed to client-owned hosting and domain where appropriate.
  • Code can be transferred to a client-owned GitHub repository, or delivered as a source-code archive on request.
  • Database/data export and setup documentation can be included where applicable.
  • Credentials and secrets are transferred securely where applicable. Required third-party services are identified before the build.
  • Post-launch support is agreed separately. The specific handover is described in the project agreement.

Scoped separately

  • Content production, migration and additional languages beyond the agreed scope.
  • Third-party accounts, recurring charges and integration limits.
  • Ongoing maintenance and later changes.

Typical technology

Technology is selected for the needs and scope; not every project uses every option.

  • OpenAI or other AI models selected for the use case
  • Server-side model calls and APIs
  • React / TypeScript for web interfaces
  • Knowledge storage and integrations when needed