Contact us
Our team would love to hear from you.
Back to cases
AI-powered learning analytics platform for children with disabilities
We partnered with Vidukate.ai to build the platform’s AI analysis engine, modernize its architecture, and improve the overall user experience.
The idea behind the project
Vidukate.ai began with a personal problem. Founder Dmitri Prilepski and his wife are raising two children with disabilities. Like many parents, they struggled to get a clear and timely picture of what was helping their children to learn, what was not working, and how to share it with everyone involved in their education.
That struggle underscores a systemic gap. In the US, parents of children with disabilities typically receive an Individualized Education Program (IEP) update from their child’s school once or twice a
year—a formal document that summarizes overall progress but rarely explains what happened during a specific lesson, how the child responded in the moment, or why one teaching approach worked better than another.
Teachers face the same problem from the other side. They often lack the time and tools to produce detailed feedback after every session.
As a result, the people closest to a child’s development may not have the information they need to make timely decisions and support progress.
Vidukate.ai was created to close that gap using AI. Rather than relying on occasional updates and fragmented feedback, the platform captures learning sessions and turns them into structured observations, session reports, and long-term behavioral patterns. This gives parents and educators a continuous, shared understanding of how each child learns, communicates, and responds.
Client
Vidukate.ai
Country
USA
Domain
Service
Solution
AI-powered learning analytics platform
Technology
Our role
Dmitri had already developed the first version of Vidukate.ai when he approached EffectiveSoft, trusting the team to treat something this personal with the same care he did. He needed a technology partner that could strengthen the existing platform, expand its capabilities, and build a technical foundation for future growth.
At the core of our work was deepening the platform’s AI-driven session analysis—the engine that turns raw learning sessions into structured, meaningful insight for parents and educators. Supporting that goal, we modernized the platform’s architecture, introduced AI-assisted practices across our own development process, and simplified authentication and account management for families, students, and education providers.
We started with a technical audit. The original codebase relied on platform-specific dependencies, which limited who on the team could contribute and made the platform harder to scale. We restructured the development environment for cross-platform compatibility so it could be built, tested, and deployed consistently across different systems, and we helped evolve the existing AWS-based microservices architecture into a more scalable foundation for future development.
Building an AI analysis system
Every learning session is recorded on video, but every lesson looks different depending on what’s being taught: a math assignment calls for a different lens than a speech therapy session.
A single, general-purpose prompt couldn’t capture that difference, so we moved the platform to subject-specific prompts tailored to different session types.
The platform’s LLM-based analysis layer adapts accordingly: for content-focused sessions, it looks at how the student completed the assignment; for student-focused sessions, it examines emotional and nonverbal responses. For combined sessions, it looks at both task completion and behavior together. This makes the resulting analysis more relevant to what was actually being taught and observed.
Across each session, the platform:
- Analyzes emotional state, motor skill development, and other behavioral aspects alongside academic performance.
- Structures and retains observations across sessions rather than treating each one in isolation.
- Generates a full report per session, including insights on what’s working well and what could be adjusted.
- Detects patterns over time—such as shifts in engagement, response to different approaches, or recurring areas of difficulty—to guide more personalized decisions going forward.
The result is a more continuous, contextual view of a student’s learning progress, built directly from what happens in each recorded session, not just an occasional summary.
AI-enabled product engineering
Sessions with AI expertsRemoving login friction
For Vidukate.ai, accessibility also means reducing the effort required to begin a learning session. The platform now uses one-time email codes as an alternative to passwords, removing friction points like forgotten credentials and lockouts. This also creates a simpler flow for people using screen readers or other assistive technology.
For recurring sessions, Kiosk Mode lets a student’s device stay paired with their account: a teacher requests pairing from the student’s device, a parent approves it once by email or in the app, and every session on that device logs the student in automatically, with no need for a repeated login step.
Enhancing account management
We also expanded the platform’s account structure so families can manage access from one place. Parent Admin Mode gives parents direct control over who can access their child’s information. They can invite teachers and education providers, assign access, review user roles, and activate or deactivate accounts without depending on another organization to make those changes.
A user who already has an account on the platform can be invited to join an organization instead of creating a duplicate account. This means a teacher working with multiple students never needs more than one login.
For students who don’t have their own email, parents can create managed accounts instead: no credentials to manage and no login screen at session time. These accounts are also what make Kiosk Mode possible for students, since there are no credentials for the device to store or enter.
How Vidukate.ai works
Education is no longer limited to traditional schooling. From classrooms and universities to corporate training and digital learning platforms, people are expected to continuously acquire new skills in a world shaped by rapid technological advancements and artificial intelligence (AI).
Outcomes
Vidukate.ai provides parents and educators with a shared, session-by-session view of a child’s learning experience and progress.
Instead of waiting for an occasional, dense IEP update, families can receive structured observations after every session. Teachers and education providers can use the same information to understand how a student responds to different assignments, environments, and educational approaches, tailor educational plans, and personalize the child’s learning journey.
Ultimately, the platform helps replace fragmented feedback and guesswork with more consistent, evidence-based educational support.
Development continues in the following key areas: improving the accuracy and relevance of AI insights generated for each session and extending the analysis so it can adapt more closely to the individual needs of each student.
Testimonial
I want to extend a sincere thank you to EffectiveSoft for joining us on this journey. As both a founder and a father, I truly appreciate a partner who understands that this isn’t just another software project; it’s about real-life situations where every detail matters.
Thank you, EffectiveSoft, for your partnership and for helping us empower families with the support they deserve.
Dmitri Prilepski
Founder & CTO at Vidukate.ai