AI from day one. Not as an afterthought.
Bluesite, Colombia's leading cloud-based psychometric testing company, had already shipped two versions of a product its clients didn't enjoy using. Molt rebuilt the platform from evidence and embedded AI into the workflow. The result: a competency assessment now takes up to 35% less time to build, and Bluesite is closing deals outside Colombia.
The essentials
- Building a competency assessment takes 20–35% less time thanks to AI suggestions inside the creation flow.
- Reports are 15–25% easier to interpret for the end user, thanks to a redesigned information layout.
- Data processing and cross-company language homologation dropped by 10–20%, accelerating B2B operations.
- Redesigned MVP shipped to production in 9 months (Jul 2022 – Apr 2023), followed by 25 months of continuous AI-powered iteration.
- Bluesite closed new business outside Colombia and grew its client base with AI-powered modules.
Two product iterations that failed to win users over
Bluesite sells a critical product: BETESA psychometric and competency assessments that companies rely on to make talent decisions. The content was solid, but the platform delivering it (already on its second iteration) had a confusing interface, a frustrating end-user experience, and usability issues that carried over from release to release.
Bluesite's team came to Molt not for an agency to execute a brief, but for a strategic partner to transform the product at the root. The pressure was twofold: they needed a new platform without disrupting clients still working on the previous version, so for the first stretch of the project, the old platform stayed accessible through a sidebar redirect.
Real partnership, not deliver-and-leave
Bluesite didn't need yet another vendor to ship a rushed third version. They needed a team willing to first understand why the previous two iterations hadn't landed, and willing to stay and iterate until the product actually worked.
That's exactly what defines Molt's model: the 80/20 rule (80% of success comes from understanding the problem correctly, only 20% is technology) and the Delta Loop methodology, combining fast cycles with AI and low-code to deliver up to 4× faster than traditional models, without creating tech dependency. The proof is in the duration: 3.5 years of continuous partnership.
Research, redesign and continuous AI-powered evolution
Molt started with research on v1 (user interviews, usage analysis and pain-point mapping) to separate what worked from what had to go. That input drove the design team, which built a new look for the platform on top of Bluesite's existing brand identity and color palette. Only then did development begin.
The project's flow stayed simple and disciplined: research → evidence-based design → development → production → continuous iteration. Once live, the next 25 months went into evolving the product: new modules, upgraded reporting with individual and group metrics, an inventory management system that gave Bluesite real control over how it monetizes each client, and AI embedded into core processes. The project didn't end on a deliverable; it ended when the product could sustain business growth.
9 months to MVP, 25 more to iterate
Three measurable savings + a business that grows
The results show up across the three fronts that matter in a B2B SaaS. For Bluesite's team, building a competency assessment takes 20–35% less time thanks to AI suggestions embedded in the flow. For their enterprise clients, data processing and cross-company language homologation dropped by 10–20%. And for the end user actually taking the assessment, reports are 15–25% easier to interpret.
The commercial change follows these numbers. Inventory management and per-case adjustable monetization gave Bluesite real control over how each client is billed. AI-powered modules unlocked expansion outside Colombia and client-base growth. The product went from holding the business back to opening it up.
| Key outcomes | Before | After |
|---|---|---|
| Time to build a competency assessment | Baseline (manual creation) | 20–35% less time with AI suggestions |
| End-user report comprehension | Reports hard to interpret | 15–25% easier with the new information layout |
| Data processing & cross-company homologation | Baseline (manual process) | 10–20% less time |
| Monetization model | Limited control over per-client pricing | Inventory management + per-case adjustable pricing |
| Commercial reach | Focused on Colombia | New deals closed abroad, client growth enabled by AI |
A product that keeps growing
With a consolidated product, AI modules in production and new markets opened, Bluesite has the foundation to keep expanding its talent offering, without depending on Molt for day-to-day operations. The partnership did what it set out to do: leave the client stronger, not tethered.
Frequently asked questions
How much time does AI save in this product?
How long does Molt take to ship a product to production?
How do I know the output will be actionable, not just a polished report?
Am I locked in with Molt after the project?
Does this work if I already have a product in production I don't want to break?
Have a product that needs to be transformed at the root?
Let's talk about how Molt can be your strategic partner, not your next vendor.
No commitment · 30-min intro call
Explore other work.
SaaS · LogisticsBuilt in iterations. Still running after seven years.
Read case study →4,000+ businesses certified across 6 states in 5 years.
Read case study →
Internal