The problem I was hired to solve
As AI coding tools evolved from autocomplete assistants toward more autonomous systems, Sonar needed more than a product roadmap — it needed a point of view on where software development was heading, and where Sonar fit in that future. Without it, the team risked building features for a workflow that was already being replaced.
What I did
- Partnered with PM and engineering leads to develop the Future of SDLC deck — a strategic document exploring how SonarQube should adapt as AI adoption moves from augmentation to near-autonomous code generation.
- Combined desk research, competitive tool benchmarking, and workflow modelling to map the trajectory of software development across three time horizons: now, next, and later.
- Synthesised findings into a shared framework that the team could use to evaluate product decisions against the direction the industry was heading — not just where it was today.
What I found
The shift underway isn't incremental. Software development is moving toward a model where developers spend significantly less time on direct implementation and more time on orchestration, specification, validation, and oversight. AI is absorbing increasing amounts of code generation, remediation, documentation, and testing.
In that environment, Sonar's value proposition changes. The question is no longer only "how do we help humans write better code?" It becomes: how do we help humans and agents understand the properties of code that was generated inside an agentic black box? — including its security, quality, architecture, performance, testability, and trustworthiness.
This reframe also surfaced a new category of jobs Sonar needs to support: helping agents iterate on analysis results, improving prompt quality, and making code quality visible inside agent-driven workflows rather than surfacing it only after the fact.
The decisions this enabled
- Created a shared strategic vocabulary for discussing AI adoption levels, workflow change, and product opportunity — across product, engineering, and leadership.
- Gave the team a concrete bridge between near-term product questions (what do we build next quarter?) and longer-term platform direction (what does Sonar need to become?).
- Helped articulate where Sonar remains indispensable as the locus of software development shifts from manual coding to agent-enabled systems — a positioning that held up in executive conversations and informed investor-facing narratives.
- Provided a framework for prioritising investment across now, next, and later — reducing the risk of optimising for a workflow that AI was already in the process of replacing.