Independent AI workflow and integration consultancy
Codance examines real operational workflows, identifies the relevant information, decisions, exceptions and risks, and shows whether AI is suitable—and, where it is, how it should be integrated.
Enterprise-grade AI integration thinking—proportionate to the decision, without enterprise theatre.
Workflow first · Vendor-neutral advice · Human authority preserved · No implementation sale required
AI projects rarely fail because nobody could find a model or automation platform. They fail because the underlying workflow, information and authority were never made clear.
Before choosing a tool, somebody needs to ask:
Codance answers those questions through focused engagements proportionate to the decision being made.
Should we proceed—and where should we begin?
For a defined workflow, Codance assesses suitability, readiness, risks and prerequisites, then recommends the most sensible first action.
You receive a professional AI Workflow Readiness Report. The conclusion may be to proceed, complete prerequisites, improve the process first, use deterministic automation, or not use AI at all.
Introductory project fee: £95
Delivery: seven calendar days after complete
requirements
How should the selected integration actually operate?
For a sufficiently understood workflow, Codance designs the future process, the roles of AI, software and people, information and integration points, approval, audit, failure handling, implementation stages and acceptance measures.
You receive an AI Integration Blueprint suitable for internal approval, supplier briefing, quotation and controlled prototype planning.
Introductory project fee: £295
Delivery: ten calendar days after complete requirements and
discovery scheduling
Final scope and any applicable tax are confirmed before work begins. Each launch package focuses on one defined workflow and one principal business function, while considering relevant dependencies and future compatibility.
Each launch engagement focuses on a defined operational workflow. This keeps the evidence, risks, decisions and recommendations specific enough to be useful, while containing the time and cost of the work.
That boundary does not mean the workflow is considered in isolation. Codance examines the systems, information and neighbouring processes upon which it depends, and considers foreseeable future needs far enough to reduce the risk of avoidable technology or data dead ends.
Where wider opportunities are identified, they are recorded as future context rather than silently added to the engagement. Connected workflows can then be addressed through a staged or separately scoped route where that is justified.
Tools, vendors and models come after the workflow, information, decisions and consequences are understood.
Deterministic software should perform deterministic work. People must retain authority for consequential judgement. AI is used where interpretation or assistance creates genuine value.
If the workflow is not ready—or AI is not the best first solution—the report will say so and explain what to do instead. Codance does not need to sell an implementation to justify its advice.
The deliverable is not an inspirational slide deck. It is a reasoned decision or solution-level design that an owner, manager, developer or supplier can act upon.
Westmoor Precision Engineering Ltd is a fictional but representative worked example. The business initially believed AI should prepare quotations automatically.
The assessment found that its immediate constraint was earlier: enquiries arrived through several routes, evidence was inconsistent, and no single record showed ownership, missing information, clarification and status. Historic estimates were not reliable enough for dependable automated pricing.
The recommended first integration became AI-assisted enquiry intake and estimator preparation—with deterministic checks and human commercial and engineering authority.
That smaller decision reduced risk, clarified the supplier brief and created a safer route toward future automation.
The questionnaire is not an administrative hurdle or a test of technical knowledge. It is how Codance learns what really happens in your organisation.
Concrete, candid answers—including uncertainty, unofficial routes and “it depends”—help us distinguish fact from assumption and produce a report that is specific enough to be useful.
Codance is led by Jonathan Arklay, whose background spans more than forty years of engineering, manufacturing, operational systems, MRP integration, application logic, data thinking and technical problem-solving.
Aurora, Codance’s internal agentic-AI reference and R&D platform, provides a practical environment for exploring persistent workflows, operational truth, model routing, retrieval, audit, human authority and recovery. It is described as internal Codance engineering work, not as an external client deployment.
If you have an operational workflow, a group of connected workflows or a real decision about AI, tell us what you are trying to understand. Codance will confirm whether the Readiness Assessment, Integration Blueprint, a staged route or a separately scoped engagement is the right fit.