For a compliance officer, the EU’s Digital Services Act reads like a dense legal text. For a software engineer, it’s a set of requirements that can be parsed, tested, and automated. Inbal Goldberger, who led policy enforcement teams at Google Trust & Safety, has seen the gap between those two perspectives firsthand. Her new startup, Codeliance, is betting that the next wave of compliance automation will be about closing it.
Founded in 2025 and based in Dublin, Codeliance is building an AI-powered platform that converts online safety regulations like the DSA and the UK’s Online Safety Act into structured obligations, test scenarios, and operational workflows [Codeliance, 2026]. The goal is to give compliance, legal, and Trust & Safety teams a system that can scan a product, gather evidence, test for compliance, identify gaps, and recommend fixes [Deloitte, 2026]. It’s a niche that sits at the intersection of legaltech, regulatory tech, and the sprawling world of platform governance.
The wedge: from legal text to test case
Codeliance’s apparent wedge is operational. Instead of just tracking legal requirements in a spreadsheet, the platform aims to use what it calls “software-engineering agents” to generate and execute compliance test cases against a customer’s own platform [LinkedIn, 2026]. Think of it as a continuous integration pipeline, but for regulatory compliance. The system is designed to understand what a rule like “mitigate systemic risks” means in the context of a specific social media feed or app store, then run automated checks to see if the platform’s controls are sufficient.
This approach targets a specific pain point. For large platforms, compliance with the DSA is not a one-time audit but an ongoing operational burden, with potential fines reaching 6% of global turnover. Manual processes are slow, expensive, and prone to human error. Codeliance’s bet is that the same automation wave that transformed security testing,shifting it left into the development cycle,is now ready to hit Trust & Safety [LinkedIn, 2026].
The team behind the code
The company is led by its two co-founders, who bring complementary backgrounds to the problem.
| Role | Name | Key Background |
|---|---|---|
| Co-founder & CEO | Inbal Goldberger | Former military intelligence officer; led global intelligence and policy enforcement teams at Google Trust & Safety, including early GDPR work [Codeliance, 2026]. |
| Co-founder & CTO | Yoni Lavi | Two decades in software development, DevOps, and product management; joined as CTO in August 2025 [Codeliance, 2026][LinkedIn, 2026]. |
Goldberger’s experience is the company’s cornerstone. She has operated at the coal face of platform policy, an arena where abstract regulations meet real-world content moderation and risk systems. Her participation in industry events, including a fireside chat with other Trust & Safety leaders, signals an effort to embed the startup within the existing community [LinkedIn, 2026]. Lavi, with a background in cognitive science and an Executive MBA, provides the technical architecture to turn that domain knowledge into a scalable product [Codeliance, 2026]. The team is small, reportedly just three employees including a QA consultant who joined in June 2026 [RocketReach, Unknown][LinkedIn, 2026].
An early-stage bet with Deloitte’s spotlight
Codeliance is pre-product and pre-revenue, operating in classic stealth startup mode. There is no verifiable public funding announcement, customer name, or partnership deal. What it does have is a slot on the stage: CEO Inbal Goldberger is scheduled to demo the platform at a Deloitte event in London focused on AI and tech-enabled compliance [LinkedIn, 2026]. For an early-stage company, that kind of spotlight from a major professional services firm is a significant traction signal, offering validation and a direct line to potential enterprise buyers.
The company says it is “currently working with early partners,” though none are named [Codeliance, 2026]. The path to revenue will depend on converting these early engagements into paid contracts, likely starting with mid-sized platforms and the compliance teams of larger tech companies navigating the DSA.
Where the wheels could come off
Every bet on a new regulatory software category carries inherent risks. For Codeliance, they are particularly pointed.
- The interpretation problem. Regulations like the DSA contain deliberately broad principles. Translating “freedom of expression” or “systemic risk” into binary pass/fail test cases is an immense technical and legal challenge. If the AI’s interpretation is too rigid, it becomes useless; if it’s too flexible, it loses its value as an objective auditor.
- The sales motion. Selling to compliance departments is notoriously slow and relationship-driven. A three-person startup without a proven enterprise sales leader faces a steep climb to land the six- and seven-figure contracts needed to build a venture-scale business.
- Reputational friction. Co-founder Yoni Lavi was identified in a news report about the cancellation of a Dublin tech event following pro-Palestinian threats, where he was described as a former occupation forces soldier [israelhayom.com, Jul 2026]. In the highly politicized realm of online safety and content moderation, founder backgrounds can become a factor in customer and partner decisions.
The company’s most plausible answer to the first two risks is Goldberger’s insider credibility. She speaks the language of both regulators and platform operators, which could accelerate trust-building. For the technical challenge, the bet is that a narrow focus on online safety,rather than all compliance,allows for a sufficiently precise rule set.
The next twelve months
The immediate milestone is clear: a successful demonstration at the Deloitte event, followed by converting interest into pilot contracts. The company will need to show it can move from a conceptual platform to a tool that generates actionable, defensible reports for real customers. Given the lack of disclosed funding, a seed round is likely on the horizon to build out the engineering and go-to-market teams.
From a unit economics perspective, the prize is substantial. A conservative estimate: if the DSA alone applies to roughly 20,000 online platforms in Europe, and even 10% of those spend an average of €25,000 annually on compliance software, that’s a €50 million annual market just for one regulation. Codeliance’s task is to capture a slice of that headache before the incumbent GRC suites or larger legaltech players decide to build the feature themselves.
The company it must beat isn’t another startup; it’s the internal spreadsheet and the army of consultants. If Goldberger and Lavi can prove that software agents are cheaper and more reliable than manual audits, they won’t just be selling a tool. They’ll be selling a new way to measure trust.
Sources
- [Codeliance, 2026] Codeliance - AI-Powered Compliance Platform | https://codeliance.com/
- [Deloitte, October 2026] Demo DAY Speakers: AI and tech enabled compliance | https://www.deloitte.com/uk/en/Industries/tmt/perspectives/demo-day-speakers.html
- [LinkedIn, 2026] Yoni Lavi | LinkedIn | https://www.linkedin.com/in/yonilavi
- [LinkedIn, 2026] Inbal Goldberger on LinkedIn: The Trust & Safety conference was so insightful | https://www.linkedin.com/posts/inbal-goldberger_the-trust-safety-conference-was-so-insightful-activity-7113994462405480449-g9AD
- [RocketReach, Unknown] Codeliance Information | https://rocketreach.co/codeliance-profile_b6457021c80d0ced
- [israelhayom.com, Jul 2026] Pro-Palestinian threats force cancellation of Dublin tech event over alleged 'Zionist spies' | https://www.israelhayom.com/2026/07/12/pro-palestinian-threats-force-cancellation-of-dublin-tech-event-over-alleged-zionist-spies/
- [LinkedIn, 2026] Jemma Levene - Center for Countering Digital Hate | LinkedIn | https://www.linkedin.com/in/jemma-levene-12938946/
- [LinkedIn, 2026] Marco Granada | LinkedIn | https://www.linkedin.com/in/marco-granada-5228a327