Echovise AI Experiment Platform

Project description

How can Echovise ethically extract actionable insights from user interactions within the AI Playground to enable proactive, personalized outreach while maintaining full GDPR compliance and user trust?

Context

Echovise is a company that helps organizations optimize their business processes using artificial intelligence (AI). To make AI more tangible for potential clients, Echovise developed the AI Experiment platform. This platform allows users to explore various AI use cases in an accessible way, supported by an integrated chatbot.

The current version of the platform has been developed by Fontys ICT students as part of a previous project. This follow-up project offers the opportunity for a new group of students to continue that work, extending the platform with additional functionality and improvements.

The existing system already demonstrates the possibilities of AI and provides initial guidance to visitors. However, the next step is to actively convert user interactions into actionable leads. Visitors explore the platform but do not yet proceed to direct contact. By capturing and analyzing their input, we can identify their interests and proactively guide them toward meaningful business conversations with Echovise.

Results

The project resulted in three key outcomes that together deliver technical, ethical, and business value for Echovise.

A fully researched and validated design for storing chat history and generating AI-based summaries only with explicit user consent, supported by role-based access control and audit logging.

Lead generation in AI systems does not have to conflict with privacy. By designing consent-first data access and limiting what administrators can see, the platform aligns business goals with GDPR principles like data minimization and transparency.

This outcome protects Echovise from legal and reputational risk while preserving user trust, which is essential for long-term platform adoption.

A documented, structured approach to working with SvelteKit, Supabase, and OpenAI in a production environment, including architectural insights, development patterns, and an implementation roadmap.

Lack of prior experience was mitigated through deliberate research methods such as document analysis, peer programming, prototyping, and benchmarking. This turned a risk into a learning asset.

This outcome reduces future development risk, improves maintainability, and provides Echovise with reusable patterns for further AI-driven features.

About the project group

We are all software engineering students. We had scheduled sprint meetings with the client in which we explain the implementation of our features. We also had a managed Jira board in which we distributed tasks equally between ourselves.