We asked our risk team to share their genuine experiences using Archetype, Jaywing’s AI risk modelling platform, day-to-day. What's different? What's better? Do any challenges remain?
For risk leaders in 2025, you need models that satisfy regulators while delivering results. You need them built quickly without sacrificing quality. And you need your talented team focused on solving problems, not wrestling with code.
In this blog, we'll take you behind the scenes with Abisola Akinjole and Nathaneal Panter to see how Archetype has changed not only what we build but also how we work.
The real benefits we're seeing
Perhaps the biggest benefit is how Archetype, as a true no-code solution, has removed technical barriers. Users don't need to learn Python or other programming languages to build sophisticated models.
"Archetype has changed how I think about modelling by making it more accessible and efficient, meaning anyone can build a model," says Abisola.
The platform's intuitive interface requires minimal to no training for users. This extends to model deployment too. Archetype can export inference (prediction) code in multiple languages including SAS, LUA, SQL, Java, Python, and PowerCurve. This means users can load their modelling data into Archetype and download ready-to-use prediction code in their preferred language without writing a single line of code.
This flexibility removes bottlenecks between modelling and implementation teams, further accelerating the delivery of value from analytics.
"The auto-select and variable selection features stood out to me," says Abisola. "I was pleasantly surprised by how quickly and efficiently Archetype handles feature engineering, something that would take considerably longer if done manually. It simplifies a complex process, making it much easier to experiment with models."
This combination of better performance and simplified workflows naturally leads to time efficiency. "I estimate that Archetype saves me about 50-60% of the time in the modelling process, maybe even more," adds Abisola.
Abisola adds a specific example: "We had to evaluate thousands of variables across a large dataset and Archetype's auto-selection, feature engineering capabilities and fast model creation made a huge difference," explains Abisola. "Tasks like feature selection, often across multiple blocks of credit bureau data, which would normally take days when done manually, were completed within hours."
But the real value is achieving superior model results with less effort.
In fact, deep learning models built in Archetype typically outperform traditional linear methods by recognising complex non-linear relationships in underlying data. This allows the platform to identify subtle patterns and interactions that conventional approaches might miss.
"The data prep and grouping" is what Nathaneal couldn't imagine working without now. For Abisola, it's "the variable specification tab. It allows for quick feature engineering, letting me set monotonicity, define rankings or groups, and easily scan through all variables in one place."
These capabilities ensure that while we use advanced AI techniques, we maintain full control over how variables interact within the model. This balance of power and explainability is particularly valuable when regulatory compliance is non-negotiable.
When asked if the platform makes it easier to produce models that satisfy documentation requirements, Abisola confirms: "Yes, I find it much easier to create governance-ready models with Archetype, as it has handy features that help with documentation and compliance, making it simple to meet regulatory requirements and collaborate better with the team."
This built-in compliance support streamlines what was previously a separate, time-consuming phase of model development.
What the team values most about Archetype
We asked our team to share the features they actually rely on day-to-day. Their answers reveal what makes Archetype truly valuable:
#1. Feature selection and engineering
"Feature selection is something I genuinely cannot imagine working without now," says Abisola. "Being able to see the feature importance, univariate importance, and correlations all in one place makes a huge difference."
This visibility into how variables interact saves hours of manual analysis. You can quickly identify which factors drive your model's predictions and eliminate redundant variables.
#2. Intuitive interface
The platform's user-friendly design came up repeatedly in our conversations with the team. "The learning curve was quite manageable, thanks to the easy-to-use interface," Abisola notes.
Nathaneal describes it as "quick and simple to understand." This accessibility means new team members can get up to speed rapidly, and occasional users don't need to relearn the system each time.
#3. Balanced automation and control
Archetype can autocreate an 80-90% model solution with its automated data cleaning, feature engineering, variable selection, and model build capabilities. This means getting a good model within an hour, allowing early evaluation of the business case. You can quickly determine whether a project is worth pursuing further or if the potential uplift doesn't justify replacing an existing model.
While automated model building is great for rapid testing, Archetype also gives modellers exceptional flexibility when developing full production models. Users can easily test different model iterations with various feature engineering approaches, variable selections, and architectural choices without coding. This balance of automation and control satisfies both those looking for quick results and those who want to fine-tune every aspect of their models.
#4. Automated data preparation
Data cleaning and preparation typically consume the lion's share of model development time. Archetype significantly changes this equation.
"The data prep and grouping" features are what Nathaneal considers indispensable. The platform handles much of the tedious work automatically, letting our team focus on analysis and interpretation.
#5. Clear visualisations for stakeholder engagement
Explaining complex models to non-technical stakeholders has always been challenging. But Archetype helps bridge this gap.
"Archetype provides clear visualisations and explanations of model outputs, making it easier to communicate complex concepts," explains Abisola. "This helps to build trust and understanding around the models we create."
#6. Project management
With multiple model builds happening simultaneously, keeping track of progress can be difficult. "Archetype makes it much easier to juggle multiple model builds or requests by allowing me to efficiently access and track different projects in one place," says Abisola.
Nathaneal values how Archetype provides a central place for all your model development projects. Including candidate and deployed models. Why? Because this allows easy access to older projects (nothing get lost) and rapid benchmarking capability when new development is on the way.
What's different about risk modelling now
Beyond the technical capabilities, Archetype has changed how we approach risk modelling as a team. Here’s a handful of the key changes we’ve seen:
From sequential to parallel working: Traditional model development can create bottlenecks. Data scientists would work on one model at a time, moving sequentially through each stage of development.
From technical barriers to collaborative modelling: Previously, modelling was limited to those with specific coding expertise. Archetype has democratised this process.
From reactive to proactive risk management: The speed of model development has helped us become far more proactive. Instead of scrambling to address issues after they emerge, we can now run multiple scenarios and anticipate challenges.
From opaque to transparent decision-making: Risk models have often been criticised as "black boxes" - particularly sophisticated AI models. Archetype tackles this challenge head-on.
|
This transparency is crucial for both internal stakeholders and external regulators. We can clearly demonstrate why a model makes specific predictions, building confidence in the results. |
From fixed to flexible models: Traditional models were often left unchanged until performance degraded significantly. The effort required to update them meant updates happened infrequently. Now, with Archetype's efficiency, we can continuously refine our models to reflect changing market conditions.
Archetype benefits for risk leadership
For leaders, the practical benefits we've discussed translate into advantages like these:
✅ Stronger audit outcomes: Regulatory scrutiny of risk models continues to intensify. Archetype's focus on explainability directly addresses this challenge.
✅ Better resource efficiency: Every risk function faces resource constraints. Archetype helps you maximise the impact of your talented team. This means your existing team can build more models, run more scenarios, and deliver more insights without expanding headcount.
✅ Faster time-to-value: Nowadays, speed matters. The ability to quickly develop and deploy accurate risk models can be the difference between seizing an opportunity and missing it entirely.
✅ Broader model utilisation: When models are easier to build and understand, they get used more widely across the organisation. This extends the value of your modelling efforts beyond traditional risk applications. For instance, Archetype is also widely used within marketing teams for predictive insights.
✅Enhanced team capabilities: Perhaps the most significant long-term benefit is how Archetype improves your team's capabilities. By removing technical barriers, it allows risk specialists to apply their domain expertise directly to modelling challenges. Archetype locks the full modelling pipeline into deployment-ready code and provides detailed reason codes for transparency. This reduces deployment time from days to hours while giving valuable insight into decision factors.
See it in action
The proof is in the platform. Our team has used Archetype to power numerous successful implementations across the financial sector.
Want to see how it works for your specific challenges? Get in touch for a demonstration tailored to your needs.
Ready to upgrade your modelling capability? Book a demo today and see how easy it can be to create better risk, fraud, and marketing models.
Contact us at [email protected] or call 0333 370 6500.