Hyperskill's CPO on Why Workforce Problems Need More Than a Training Catalog

Skills matrices in spreadsheets, development plans in documents that go stale within a quarter. Every organization doing annual planning lives with some version of this, and the tools that exist were built to handle each piece separately.

The idea for Aprena (https://aprena.pro/) came from the management side: years of building grade matrices and running calibration cycles, trying to turn evaluation results into plans people could actually act on. Its founder is the CPO of Hyperskill, with 12 years in educational technology behind her.

The tools in this space are fragmented. Skills matrix software, performance evaluation platforms, learning management systems: each does its job in isolation, with no meaningful connection between them. Nobody has built the connective tissue, and that's the gap Aprena is trying to fill.

‍

First steps in the product development

The original version of the product was focused on career orientation for mid-career specialists. The goal was to show people where they stood in the market and what paths were open to them. While building it, a series of interviews with HRBPs and people strategy leaders revealed how companies approach that same question from the inside.

What came back was the same problem from the other direction: companies didn't know how to evaluate what their people actually knew, or how to calibrate those assessments fairly across managers. The pain was real and recurring, and clearly larger than what the original product addressed. That was the pivot.

‍

What the product does

Aprena is for HRBPs, people strategy specialists and CEOs, specifically in moments of significant change: a restructuring, a strategic shift, or planning cycles where the budget is tight and the list of deliverables is long.

The core question it helps answer: given what the company needs to achieve and what it has to work with, what is the most efficient way to cover the required skills? The answer might include learning, but only for employees who could realistically move into a different role. For other gaps, it might be a contractor, an AI tool, or a different team structure. Aprena puts all of those options in one place.

The most concrete early result came from a startup mid-fundraise, with a small team and an ambitious product plan. Aprena helped them build a workforce strategy that accounted for what AI tools could genuinely handle and identified where contractors made more sense than full-time hires. The outcome was a plan that let them avoid hiring roughly 10 people they would otherwise have needed to reach a real launch.

‍

How it was built

Two MVPs so far. The first, focused on career assessment, took a month and a half with one backend developer. The second, after the pivot, took two months with an AI engineer. A third iteration is in progress.

What was deliberately left out at each stage: anything visual, anything beyond the core workflows. The goal was a system that could walk through its key scenarios end-to-end. Integrations with external services are planned but pushed to the next iteration; they're not what you use to validate a product.

The technical approach is built around context management. The system prepares the right information for each step: skill profiles, roadmap data, previous decisions. This keeps the reasoning focused and outputs consistent across long workflows. Persistent memory means each engagement builds on the last. The stack runs on Python, Node.js, PostgreSQL, Qdrant for vector storage and OpenAI's current models.

‍

What lessons were learnt

Even with 12 years of product experience, first-time founders hit the same wall. The product decisions weren't the hard part. Keeping going was: when the first version doesn't land, when early customers take real effort to convince.

The advice is to find the single most painful scenario for your user. A real problem they have today. Solve that completely before anything else. Hold a longer-term picture of where the product could go, but don't let it dilute the one thing that makes someone actually pay for what you've built.

Twelve years of edtech expertise isn't something you can shortcut, but the engineering layer this product is built on is learnable. It's the same stack that makes products like Aprena possible to ship with a small team. If you want to build at that level, the AI Engineering Bootcamp is where to start.

Skills gaps usually end with a training recommendation. We've built a tool that works out whether training is actually the right call.