From Skeptic to Systems Builder
In 2015, Volodymyr Radko was a student in Poland looking for his first real break in tech. With a background in Pascal and C#, he was close to accepting a .NET developer role, when a recruiter mentioned Salesforce instead. After a three-month trial period, he landed his first job as a Junior Salesforce Developer and never looked back. Today, he's a Salesforce Architect and Lead Developer at Noltic, where his personal motto still guides him: "Never stop learning."
A More Honest Relationship with AI
Like many developers, Volodymyr was an early AI adopter, using the technology to speed up routine tasks. But over time, his perspective shifted. He stopped thinking of AI as 'magic in a prompt' and started treating it like any other part of a system - something you have to set up properly, give the right boundaries, and connect to real data and real interfaces. Otherwise it never gets past a demo.
That shift changed how he built with AI. It inspired two open-source projects focused on a challenge many organizations face: moving AI from impressive demos to reliable enterprise systems. SF Agent Bridge connects AI agents directly to enterprise software, while SF Compact helps make those agents more dependable in production.
Building the Agentic Enterprise, One Workflow at a Time
Volodymyr's day-to-day setup reflects his disciplined, systems-first approach. Rather than relying on a single chat assistant, he runs a connected chain: Claude Code paired with MCP and the Salesforce CLI . The agent has access to tools, the command line, and the full project structure, and it works within well-defined boundaries. The payoff is real: faster work, less routine. But it comes from how you set things up around the agent, not just from the fact that “we have AI now”.
Today, he's focused on building workflows where AI agents can work together reliably, automating routine processes while keeping people in control.
It's great to see modern technology trends coming to Salesforce. It gives me confidence that I'm heading into the future, not into the past.
Volodymyr RadkoSalesforce Developer, Noltic
What Comes Next
What excites Volodymyr most right now is not the next smarter model. It is the tooling layer maturing around AI: MCP, CLIs, and structured contracts that let agents act inside real systems rather than just answer questions in a chat window.
His advice to fellow Trailblazers is simple. Learn the fundamentals. The value of writing code by hand is dropping fast, but the value of understanding systems is rising. You need to know how the pieces work, because without that, you simply cannot verify what the AI produces.