Building Production-Grade Retrieval Augmented Generation (RAG) Systems
We developed an interactive learning environment to train engineering teams at a global organisation in RAG systems, ensuring comprehensive technical understanding.
Success Stories
Evaluating AI Maturity Across 200+ Financial Specialists in 8 Weeks
Our client had already invested heavily in AI tools, foundations training, data infrastructure and practical use-case delivery. The next challenge was not simply adoption, but maturity: understanding how close the organisation was to leading peer firms, where value was already emerging and what was blocking broader, governed, confident use of AI across the business.
Client
Global bank
Industry
Financial Services
Specialism
AI
The challenge
The client was not starting from a standing position. Senior leadership had already created a clear top-down push for AI adoption and had backed that ambition with meaningful investment: enterprise access to AI tools, foundation training, a strong technology and data stack, external support for AI use-case builds, and a growing set of live AI use cases, including agents.
That progress created a more sophisticated challenge. Leadership did not just need to know whether people were using AI. They needed to understand whether the organisation was becoming truly AI mature, how capability compared with leading peer firms in the sector, and where the business was still leaving value on the table.
In practice, this meant moving beyond a narrow training-needs analysis. The client needed a credible baseline across every employee and function, showing where capability was strong, where confidence was uneven, which behaviours were becoming embedded, and which cultural, governance, data or workflow barriers were preventing teams from realising more value from AI.
What made the challenge difficult
- AI activity was already happening at pace
The client had multiple initiatives in flight, including training, tooling, external delivery support, use-case experimentation and agent-based solutions. The issue was no longer activity; it was visibility, consistency and maturity. - Maturity had to be measured role by role
A generic questionnaire would not have been credible. Employees in technology, operations, investment, finance, compliance, HR and leadership roles interact with AI in different ways and need to be assessed against different behaviours. - The output had to support business change, not just course design
Training was one outcome, but the broader value was in identifying population-wide trends, mindset gaps, cultural blockers, governance needs and foundational changes required to make AI valuable at scale.
- The process had to be lean
A large consultancy diagnostic could have taken months and created heavy overhead. The client needed something rigorous enough for leadership decisions but practical enough to move quickly.
Our approach
Neueda designed a lean, evidence-led AI maturity baseline that combined leadership discovery, role-tailored assessment, full-population interviews and structured analysis. The engagement was designed to establish a defensible view of current maturity while also giving the client a practical activation plan for what should happen next.
The assessment was built around smart questionnaires tailored to functional roles. Questions were designed to test consistency through triangulation rather than relying on simple self-reported confidence. This made the data more useful because it surfaced how people actually approached AI in realistic scenarios, not just how comfortable they felt using the tools.
Every employee was assessed and interviewed, rather than relying on a small sample. This created a detailed maturity picture at individual, role, functional and organisational level, allowing leadership to identify patterns across the whole population and make decisions based on evidence rather than anecdote.
Delivery at pace
In under two months, the team completed leadership discovery, assessment design and build, rollout, full-population interviews, analysis and recommendations. The approach compressed what would often become a much heavier consultancy exercise into a focused, practical engagement that still produced the evidence needed for leadership decision-making.
| Step | Workstream | Purpose |
| 1 | Leadership discovery | Aligned the maturity ambition, business context, existing AI investments and the change outcomes leadership wanted to drive. |
| 2 | Assessment design | Built role-tailored questionnaires with triangulated questions to check consistency and assess practical behaviours. |
| 3 | Full-population rollout | Captured input across the whole employee population, avoiding the limitations of a representative sample. |
| 4 | Interviews | Interviewed every employee to deepen the evidence base and understand real-world confidence, blockers and value opportunities. |
| 5 | Analysis and activation planning | Identified trends, themes, barriers and recommendations across capability, culture, governance, communities, vendor pathways and data foundations. |
Recommendations across the full AI maturity system
The output was deliberately broader than a training plan. The analysis created a set of recommendations across the conditions needed for AI maturity: mindset, structured capability build, peer learning, vendor pathways, prototype support, governance, data access and the route from experimentation to production.
| Workstream | Primary owner |
| Awareness, culture and mindset | Client owns, with Neueda advising |
| Structured capability build | Neueda owns |
| Communities, champions and peer learning | Client owns, with Neueda designing and seeding |
| Self-service and vendor pathways | Client owns, spanning internal platforms and vendors including Cornerstone, Microsoft, Snowflake, FactSet and Bloomberg |
| Delivery partnership and prototype support | Client and implementation vendor own |
| Governance, policy, approvals and route to production | Client owns, with Neueda input on standards and learning wrapped around them |
| Data foundations and access | Client owns |
Results and impact
The engagement gave the client a credible, organisation-wide AI maturity baseline that could be used for far more than training design. It showed where capability already existed, where confidence was ahead of practice, where practical use was being constrained, and where leadership needed to intervene to unlock value.
Because every employee was assessed and interviewed, the output gave leadership a detailed view of trends and themes across the whole population. This helped separate isolated examples of AI activity from repeatable patterns of maturity, and gave the business a clearer line of sight from individual capability to enterprise-wide change.
The recommendations gave the client a practical route from baseline to activation: targeted capability build where Neueda could lead, communities and champions to embed peer learning, clearer self-service and vendor pathways, stronger governance around the route to production, and attention to the data foundations that underpin scalable AI use.
Most importantly, the work helped shift the conversation from ‘how do we train people on AI?’ to ‘what needs to change across the business for AI to create meaningful, governed and repeatable value?’ That broader view is what turns AI adoption into AI maturity.
Get in touch
Could your organisation benefit from an AI maturity baseline? Set up a call with our experts today.
Related Success Stories
Building Scalable AI Agents at a Global Finance Company
When our financial services client decided to scale their AI solutions, we delivered immersive, practical training to their engineering teams in developing and implementing multi-agent systems.
Empowering 200+ Leaders to Make AI-Enabled Decisions
Our training empowered senior leaders at a global organisation to make smarter, AI-driven decisions through targeted, experiential skills training.