Agent
One agent handling one clearly defined process.
What we define together
- Inputs, tools and the expected output
- Permissions and points requiring human approval
- Action records and exception handling
Q-Dev / Agentic
Agents carry out agreed tasks, use approved sources and act within defined permissions. We match the setup to your process: one agent, a team of agents, answers from company documents or voice interaction.
Connected to your work
Voice requires a website and Domain or a suitable knowledge base. We can use your existing website and materials you have the rights to use. If those materials are ready to use, you do not need a separate Data service. We also agree when the agent should pass a question to a person because its sources do not support an answer.
From agreement to deployment
Every Q-Dev project has four stages with an approval meeting after each one. We document the agreed decisions and revisions together.
We define the process, scope, solution design and acceptance criteria.
We build the first working version and check the core scenarios.
We launch the agreed solution in its target environment.
We prepare your team to use the solution and start maintenance.
The document following each gate sets out the approved number of revisions. Once approved, you have two additional rounds of changes to the completed work for that gate.
Timelines are agreed for each project. Changes may affect the schedule; the price changes only when work goes beyond the contract scope.
Rights to the code transfer to you after the entire project has been paid for.
Q-Journey
When an application follows an audit and is designed to improve your organisation, we deliver it in stage two of Q-Journey — including applications that do not need AI.
Let’s start with you
Tell us who does the work today, how it happens and what needs to change. We will agree on the next step together and prepare an individual proposal.
Describe one process We prepare a quote after discussing and defining the scope.