ITACC Labs
Where our next ideas are built and tested
ITACC Labs is our research and innovation practice. It keeps our client work ahead of the curve, and it's where our own products come from.
Research areas
What we're working on
We focus on the questions that decide whether AI can be trusted in production.
AI agents & tool use
How AI systems can plan and act in real software safely, with the right permissions and human checkpoints.
Retrieval & knowledge systems
Grounding answers in trusted sources, with citations, freshness and access control.
AI security & safety
Finding the security gaps that appear when AI writes code and configures systems. Our first product, Rowlock, came from this work.
Evaluation & reliability
Methods to measure accuracy, consistency and cost, so AI systems can be trusted in production.
Our first product
Rowlock
AI coding assistants make it possible to ship an app in days. They also make it easy to ship database access rules that expose more than intended. Our research into apps built this way led to Rowlock: security reviews of database rules and configuration, as one-time audits or continuous monitoring with Rowlock Watch.
- Reviews of database security rules
- Plain-language reports with fixes
- Monitoring of every new change
From lab to product
How an idea becomes something you can use
Stage 1
Explore
We study emerging AI capabilities and the problems they create for real teams.
Stage 2
Experiment
We build prototypes, test them on real-world examples and measure the results.
Stage 3
Productise
Ideas that prove their value become products and reusable components for clients.
Have a problem worth researching?
We partner with organisations on applied research and early products. Tell us what you're exploring.