Developer launches Ekselio finance workflow builder on Hacker News

By Splitfeed EditorialLast updated Sep 26, 2026, 18:55 UTC

On September 25, 2026, a developer identified as KD posted Ekselio to Hacker News under the banner of a Show HN submission, describing it as a local-first AI-assisted builder aimed at finance workflows. KD explained a career path that began with basic web skills in college before two decades in finance and mergers and acquisitions took over. During the COVID-19 pandemic, KD returned to coding, picking up React and TypeScript, and the subsequent emergence of tools like ChatGPT and Claude provided the final push to build Ekselio as a side project. The pitch frames the tool as a domain-specific answer to general AI builders, with Lovable cited as a direct inspiration.

The significance of this launch sits less in the tool itself and more in what it represents as a pattern. Finance workflows, particularly in M&A, involve highly sensitive data, complex logic, and regulated processes, all of which create real friction when using generic AI builders. A local-first architecture directly addresses the data-residency concern that would otherwise block adoption in many financial contexts. Whether a solo side project can match the reliability and depth that professional finance teams expect is the central question the product will face, and that question is genuinely open at this stage.

None of the three camps have posted reactions to this specific story yet. Pro-AI commentators would typically welcome a domain expert applying AI tooling to a specialized field, arguing that this is exactly the kind of targeted, experienced-led development that produces durable products. Anti-AI voices would likely raise concerns about AI-generated outputs entering high-stakes financial decisions, where errors carry legal and monetary consequences that a side project may not be equipped to handle. Middle-ground observers would probably focus on the local-first design as a sensible concession to legitimate data-privacy concerns, while reserving judgment on whether the AI assistance layer adds genuine value or introduces new risks.

The clearest signal to watch for is early user feedback within the Hacker News thread itself, where finance professionals often surface specific workflow requirements or security objections that stress-test a product's real-world fit. Any follow-up post from KD describing traction, a public demo, or a formal release would also clarify how far the project has moved beyond its side-project origins.