ClearMetrics

Turning complex skill data into decisions people can actually act on

The core challenge was not collecting skill data. It was making sense of it in a way that felt personal and actionable rather than generic and discouraging. A traditional assessment spits out a number. ClearMetrics needed to take that number, compare it against a specific target role, identify exactly which skills were strong and which were holding someone back, and then tell them precisely what to do about it.

The AI layer had to work across a wide range of roles and industries without losing specificity. A recommendation that works for a junior content writer targeting a content strategist role needs to be very different from one built for a data analyst targeting a product management position. Generic advice would have made the platform feel like every other assessment tool on the market.

The platform also needed to sustain engagement over time, not just deliver a one-time result. Skill development happens over weeks and months. The product needed tracking, retake prompts, progress visualisation, and a way for managers to review their team's results without the experience feeling like surveillance.

A complete AI assessment platform built around career intelligence

We built ClearMetrics end to end using React.js for the frontend, with a Python-powered AI backend that handles the assessment analysis, role gap mapping, and learning recommendation generation. The assessment engine is built around a role-specific competency framework that maps each skill dimension to the requirements of hundreds of target roles across different industries.

The AI analysis layer uses OpenAI at its core but with a significant amount of custom prompt engineering and context management built on top. It takes a user's raw assessment scores, the target role's competency benchmarks, and their historical assessment data if available, then generates the kind of specific, nuanced feedback that actually helps someone understand what to prioritise.

The manager-facing layer lets team leads review their direct reports' assessment results, identify skill gaps at a team level, and share specific recommendations without the individual user feeling micromanaged. It was one of the features that made ClearMetrics genuinely useful at an organisational level, not just for individual career development.

Built for nuanced AI analysis and real-time data visualisation

React.js handled the frontend with a component architecture built for the complex data visualisation the dashboard required. Node.js and Express managed the API layer. Python powered the AI analysis backend, including the role gap engine and recommendation logic. OpenAI API provided the language model foundation with custom prompt layers on top. MongoDB stored assessment records, user profiles, and longitudinal skill tracking data. AWS managed hosting and compute. Twilio handled notification and assessment reminder workflows.

Users who finally understood their own skill gaps clearly

The most consistent piece of feedback from early ClearMetrics users was that it was the first assessment tool that made them feel understood rather than judged. The specificity of the AI analysis, the clarity of the role gap breakdown, and the directness of the next-step recommendations gave people something they could actually use rather than a score they did not know what to do with.

Completion rates significantly outperformed what the client had seen from traditional assessment tools. Users were finishing the full assessment flow and coming back for retakes, which had historically been the hardest behaviour to drive in this category. The AI personalisation was the factor that made the difference.