Tools that developers
actually love
Developer tooling is about more than features. It is about understanding how engineers think, work, and solve problems, then building tools that fit seamlessly into that reality.
Great tools disappear
into the workflow.
The best developer tools are invisible. They do not demand attention or context switches. They surface the right information at the right moment, in the right place, and then get out of the way.
This principle has guided every developer tool I have built. At Sonar, I lead developer tooling strategy across the CLI, IDE extensions, and the platform integrations that connect code quality analysis to the tools developers already use. The goal is never to add another dashboard or another tool to check. It is to embed intelligence directly where developers work: in their editor, in their terminal, in their CI pipeline.
At Snyk, I experienced firsthand how developer experience drives adoption. The developer security tools I helped build succeeded because they met developers where they were. IDE plugins that showed vulnerabilities inline. CLI tools that fit into existing scripts. Pull request checks that provided actionable context, not just red flags. This approach was core to reaching over seven million developers.
Building platforms that scale with engineering organizations
Platform engineering is about creating the foundations that allow product engineering teams to move faster without sacrificing quality. My experience spans the full lifecycle of platform tools, from initial architecture to enterprise deployment.
IDE Integration & Extensions
Deep integration into the editor is where developer tools create the most value. I have led the development of IDE extensions that surface code quality findings, security alerts, and AI-generated suggestions directly in the editor. The key challenge is not the integration itself but the UX: how do you present complex analysis results without overwhelming the developer? The answer lies in progressive disclosure, smart prioritization, and respecting the developer's flow state.
CLI Tools & Automation
Command-line tools remain the backbone of developer workflows, from local development to CI/CD pipelines. At Sonar, I reintroduced the CLI as an agent-first tool, rethinking what a command-line interface means when both humans and AI agents are users. This required structured output formats, deterministic behavior guarantees, and composable commands that work equally well in shell scripts, CI pipelines, and AI agent orchestration systems.
CI/CD & DevSecOps Integration
Shifting quality and security left means embedding checks into every stage of the development lifecycle. I have built integrations that analyze code at commit time, during pull request review, and in deployment pipelines. The challenge is balancing thoroughness with speed: developers will bypass tools that slow them down, no matter how valuable the analysis. My approach combines fast incremental analysis for developer feedback with comprehensive scanning for quality gates.
Enterprise Platform Architecture
Enterprise platform engineering adds layers of complexity: multi-tenant architecture, role-based access control, audit logging, compliance reporting, and integration with existing corporate infrastructure. At Sonar and Snyk, I have navigated these requirements while maintaining developer experience. At LEAD Energy, I built an IoT platform on AWS serving over 100,000 smart devices, giving me deep experience with distributed systems, real-time data pipelines, and infrastructure that must never go down.
Product-led growth through genuine developer value
At Snyk, I pioneered a product-led growth motion built entirely on developer education. The strategy was simple but powerful: teach developers about security, earn their trust, and let the product sell itself. The execution was anything but simple.
I led the creation of Snyk Learn, a security education platform that taught developers about vulnerabilities, secure coding practices, and threat modeling. The content was not marketing material disguised as education. It was genuinely useful, technically rigorous training that developers sought out because it made them better at their jobs.
The results spoke for themselves. The educational content consistently ranked in the top three search results for developer security topics, driving enormous organic traffic. This reached over seven million developers worldwide and became one of Snyk's most effective acquisition channels. More importantly, it built lasting brand affinity with the developer community.
This experience taught me that the most effective PLG motion for developer tools is genuine education. Developers can immediately tell the difference between content designed to help them and content designed to sell to them. Investing in the former pays dividends that the latter never can.
Quality as a platform, not a gate
Code quality analysis has traditionally been positioned as a gate: a check that happens at the end of the process, blocking merges and frustrating developers. I believe it should be a platform, an always-available service integrated into every stage of development.
At Sonar, I am building toward this vision. The combination of IDE-level analysis, CLI-driven local checks, and CI/CD pipeline integration creates a quality platform that developers interact with continuously, not just at merge time. Adding AI into this equation transforms it further: instead of just identifying issues, the platform can suggest fixes, explain root causes, and even remediate problems autonomously through AI agents.
The integration of AI with static analysis is particularly powerful. Static analysis tools like SonarQube have decades of accumulated knowledge about code patterns, vulnerabilities, and best practices. LLMs bring natural language understanding and code generation capabilities. Combining them creates systems that can both find issues with precision and fix them with contextual awareness. This is the foundation of the next generation of developer tooling.
My work at Snyk proved this approach works in the security domain. The AI-powered security analysis engine I helped build combined symbolic analysis (traditional rule-based scanning) with neural AI (machine learning models trained on vulnerability patterns). The hybrid approach delivered accuracy that neither approach could achieve alone, and it set the template for how I think about combining AI with deterministic analysis tools.
Building developer tools
that matter?
From platform engineering strategy to developer experience design to PLG growth motions, I can help you build tools developers genuinely want to use.