Trusted in High-Stakes Assessments: AI supports. Humans decide.
Qpercom integrates AI into our digital platform to enhance assessment design and actionable feedback in interviews, observational assessments, knowledge tests and portfolio activities. As we integrate, we always do so with the goal of keeping human oversight at the core of what our platform offers.
Our Commitment: Human in the Loop
We believe that AI should not replace anybody but rather support the various stakeholders involved in all assessments. Qpercom is actively looking to add AI support for administrators when authoring questions and scoresheets and synthesising actionable feedback, for assessors so they can focus more on objective scoring and rating while providing accurate feedback and finally for candidates, to get the most from their learning journey.
Humans i.e. data controllers always make the final decision. AI enhances feedback, scoring insights, and efficiency. Results and outcomes remain transparent, reliable, and defensible.
EU AI Act Compliance
Qpercom recognises our platform’s classification as a high-risk AI system under the EU AI Act. We meet this challenge through:
- Continuous compliance with EU AI Act requirements and updates
- Comprehensive risk management across all AI processes
- Robust data governance to ensure data integrity, security, and quality
- Transparency in documenting AI’s role and limitations within our platform
- Human oversight, with AI tools designed for supervision, intervention, and review
- Regular training and awareness programmes for Qpercom staff on responsible AI use
Collaborative AI Research
We actively explore AI innovations that improve assessments, while ensuring data control remains with our clients. Our research focuses on generating structured, context-aware learner feedback and enhancing psychometric tools and scoring reliability.
As data controllers, Qpercom clients retain full intellectual property rights over all AI-generated outputs. Qpercom remains a data processor, ensuring GDPR and EU AI Act compliance.
Our Position
Qpercom ensures any integrated AI is:
- supportive, not decisive
- transparent, not a black box
- aligned with human judgment, not replacing it
From interviews to observational assessments, knowledge assessments to portfolio entries and face-to-face assessment to online remote assessments, Qpercom guarantees that AI we integrate into our digital platform supports people in completing assessments but humans always decide.
Our AI Integrations
Qpercom currently integrates with both AWS Bedrock and OpenAI, giving us access to multiple AI models. Our platform also supports the Model Context Protocol (MCP), an open standard that allows AI models to connect securely with other tools and data sources. Every AI-supported workflow is designed so that the user stays in the loop and gives final sign-off.
Our AI Roadmap
The next stage of our roadmap is a shift from Large Language Models (LLMs) towards Small Language Models (SLMs). SLMs are more efficient and require far less computing power and energy, giving them a significantly lower environmental impact. Because they can be tailored to the specific tasks within our platform, they also deliver faster response times, more consistent and predictable outputs, and lower operating costs. Their smaller footprint makes it easier to host them within controlled environments, giving us greater control over data privacy, security and data residency.
