pre-orders
accelerated user adoption
delivery timeframe met the programme
About the company
Croner is the leading provider of expert intelligence resources for tax, audit, accounting, and HR professionals. The business digitises legislation and regulatory guidance and enriches it with proprietary commentary from in-house subject matter experts.
This combination of structured content and specialist insight is Croner’s core IP, serving customers from independent tax advisors to large corporates and government bodies including HMRC.
Croner is owned by Peninsula Group, a global technology-enabled business offering innovative employment law, HR and health and safety services.
The Challenge
The challenge
As AI adoption accelerated across the professional services industry, Croner identified a clear opportunity to transform how customers access and apply their content, offering enriched experiences like never before. Rather than navigating extensive documentation, an AI-powered assistant could deliver immediate, precise guidance and insights, unlocking significantly more value from Croner’s proprietary library. With customers increasingly expecting intelligent self-serve tools in their workflows, the business made a strategic decision to move quickly and capture the market.
The foundations for this were already in place; in a prior collaboration, Codurance had architected a Databricks-powered content delivery platform for Croner, structuring and standardising the tax content that would become the backbone of the AI product. This investment meant the team could move directly to building the AI assistant rather than resolving data quality issues, significantly compressing the delivery timeline. Croner chose to partner with Codurance again to bring the engineering rigour, whilst supporting an AI-first approach to deliver a production-grade product at pace.
The Project
What we delivered
Codurance collaborated with Croner’s inhouse team over a six month period to deliver the new AI product “Croner Intelligence”, a Retrieval Augmented Generation (RAG) virtual assistant built on Amazon Bedrock and powered by AI, enabling customers to get precise, cited answers from Croner’s content library in seconds.
Alongside the core product, the team built evaluation pipelines to continuously benchmark answer quality, ensuring the assistant met a rigorous standard before and after launch. Strict guardrails and system prompts kept responses within guidance rather than advice, addressing regulatory sensitivities from the outset.
LLM observability tooling provided real-time visibility into how the assistant reasoned, enabling ongoing tuning in production, whilst a business metrics dashboard gave Croner immediate financial control over token consumption and per-user cost.
Codurance experts applied our AI-first software craftsmanship approach to accelerate software delivery while maintaining quality, control and confidence through robust engineering practices and quality gates.
The handover included a comprehensive AI skills library, developed by the team, leaving Croner fully equipped to enhance and support the product independently.
The Outcome
Outcomes
Croner Intelligence was successfully launched on budget and schedule. The commercial results exceeded all pre-launch projections:
- Significantly exceeded budgeted pre-orders.
- Achieved almost 2× the projected user adoption within the first 10 days.
- New AI product offered as a bolt-on to existing subscriptions at contract level, removing friction to access.
- Delivered accelerated ROI, and exceeded the market share projected in the initial business plan.
- Additional content products planned in the following quarters, expanding coverage.
- Clear pathway for group-wide rollout across Peninsula businesses.
Julien Gruhier
Director of AI, Croner
Alistair Brown
Global AI & Technology Director, Croner