Case study

Building full features with a small team: ADLC at uQualio

uQualio is a Danish video eLearning platform (EdTech SaaS). Like most vertical SaaS companies, uQualio has a roadmap that is always bigger than its engineering team. Features waited, requirements got discovered mid-build — and every senior hour matters.

In 2026, uQualio decided to try a different approach: ADLC — an AI-assisted development workflow built by Clever Spark. Not another coding assistant, but a workflow that covers the whole path from ticket to pull request, with a human in control at every critical step.

What we did

We did not install a tool. We built the workflow around uQualio's own project: their stack, their conventions, their way of working. The scope was agreed up front, at a fixed price, with a clear go/no-go structure.

The workflow covers:

Refinement — the AI analyzes the ticket and the linked documents from the team's repositories, asks questions, and makes implicit assumptions explicit — before anyone writes code

Planning and build — the feature is developed with the AI, following the team's conventions

Review gates — a human owns the decision at fixed control points; nothing merges on the tool's judgment alone

What changed

Full features in one leap. Work that would normally be split into two to five tickets now often ships as one coherent block. Example: a complete magic-link feature was built in one pass — and the core is reusable for the next part of the product.

Hidden questions surface early. In Christian's words, the refinement stage is "really what is lifting it": it finds the open questions that normally appear days later, in the middle of building.

The team stays in control. Two work items in parallel proved to be the practical optimum per person — the workflow raises throughput, but the human review capacity sets the pace. The workflow does the heavy lifting; the team decides where it adds the most value.

In the customer's words

"It has made it possible to create a full feature without having to necessarily understand all the things before you start refining and building."

— Christian Bjerre Nielsen, Co-Founder & Chief Product Officer (CPO), uQualio

What we learned (honestly)

This project was different from the ones where we had implemented the workflow before — a new codebase, a new domain, an agreed fixed scope. The main lesson: adapt the workflow to the existing project, not the project to the workflow. That lesson is now built into how we start every new implementation.

What's next

The agreed scope is delivered and the team uses the workflow in daily work. Now we measure.

ADLC ships with a metrics dashboard, so the results are not a matter of opinion. It tracks:

work in progress

throughput

cycle time

adoption

quality

Over the next 2–3 months we are tracking these numbers together with uQualio. What we expect: the team ships more tasks and throughput goes up — with one shared rule: quality stays the same or better. The results will extend this case study.

Clever Spark builds AI development workflows (ticket → PR) for B2B SaaS teams — designed around each team's own stack and conventions, with humans in the loop at every critical step. In production at a US insurance-software company since February 2026.

Does your roadmap grow faster than your team? Talk to us →