AI-Powered Academic Admissions: From Documents to Enrollment
- Customer
- MULTIVIX
- Project manager on the customer side
-
- IT Provider
- RUBEUS and Intelliway
- Year of project completion
- 2025
- Project timeline
- March, 2024 - June, 2025
- Project scope
- 2000 man-hours
- Goals
- The project aims to revolutionize the student admissions experience through artificial intelligence and automation. By applying OCR, Natural Language Processing, and Generative AI, the system automatically analyzes academic transcripts, matches course equivalencies, and generates digital enrollment offers. The main objectives are to digitize and accelerate the admissions process, reducing manual effort and processing time from days to hours, while ensuring accuracy and consistency in academic evaluations. Integrated with CRM and ERP systems, the solution enhances efficiency, transparency, and engagement across all admission stages, creating a seamless and intelligent journey for future Multivix students.
- Project Results
- The project delivered measurable improvements in efficiency, accuracy, and student experience. Analysis time was reduced from 7 days to less than 2 hours, while manual effort dropped by 95% (from 248 to 10 clicks per evaluation). The AI engine achieved 92.3% accuracy in course equivalence matching, ensuring consistent academic decisions. The process now runs fully digital, with contracts, tuition calculations, and enrollment automated through API integrations. The admissions team was optimized from 22 to 14 members, maintaining the same SLA. The pilot for second-degree admissions validated scalability, and the solution is now being expanded to distance-learning transfers, representing 50% of all academic analyses.
The uniqueness of the project
The project stands out for applying Artificial Intelligence to one of the most complex and human-centric processes in higher education: academic admissions. Unlike traditional systems, it combines OCR, NLP, and Generative AI to interpret transcripts, identify course equivalencies, and automate enrollment decisions. The solution achieves over 92% accuracy, transforming a manual, error-prone process into a fully digital and intelligent workflow. Its seamless integration with CRM and ERP systems enables real-time communication, financial automation, and end-to-end tracking. This unique blend of AI, automation, and academic insight makes it a pioneering model for digital transformation in education.- Used software
- The solution was built using a modular, cloud-based architecture with scalable microservices. The backend was developed in Node.js (v18) with Express.js, integrated with Python 3.11 AI modules for OCR and Natural Language Processing using spaCy, Transformers, and LangChain. OCR processing employs Tesseract.js and OpenCV for data extraction. Databases include MongoDB, PostgreSQL, and a FAISS Vector Store for semantic embeddings. The infrastructure runs in Docker containers orchestrated by Kubernetes, with asynchronous communication through RabbitMQ and continuous integration pipelines in Azure DevOps. The system integrates securely via RESTful APIs with the Multivix Rubeus and TOTVS RM platforms.
- Difficulty of implementation
- The project presented significant technical and organizational challenges. Integrating AI into a complex academic evaluation process required building a hybrid model combining OCR, NLP, and Generative AI with over 92% accuracy. Data variability across transcripts from multiple institutions demanded extensive normalization and training of semantic embeddings. Ensuring secure, real-time integration between the AI engine, CRM, and ERP systems required robust API architecture and data governance. Change management was also critical—training analysts and aligning academic coordinators with the new AI-driven workflow. Despite these challenges, the agile approach and close collaboration between IT, academic, and operations teams ensured a successful rollout.
- Project Description
-
The AI-Powered Academic Admissions: From Documents to Enrollment project reimagines the student admissions experience through Artificial Intelligence and intelligent automation. Developed by Multivix in partnership with Intelliway, the solution automates the complex process of academic transcript evaluation for student transfers, second-degree admissions, and re-enrollments.
Traditionally, this process was entirely manual — analysts downloaded documents, compared courses one by one, generated proposals, exchanged emails with students, and manually entered data into the ERP system. This caused long response times, high operational costs, and inconsistent evaluations.
The new system uses an integrated AI workflow that combines Optical Character Recognition (OCR), Natural Language Processing (NLP), and Generative AI to read academic transcripts, identify course equivalencies, and automatically suggest optimal curriculum pathways. The AI model ensemble achieves over 92% accuracy, processing each transcript in under 10 seconds.
A front-end interface enables analysts to review and approve AI recommendations in just 10 clicks — a 95% reduction in manual effort. Once approved, the system automatically generates the contract, applies scholarships, calculates tuition fees, and updates the ERP through secure REST APIs. The entire student journey — from document submission to enrollment — occurs digitally in the Multivix admissions portal.
The solution also integrates with the CRM, triggering personalized communications via email, WhatsApp, and call center follow-ups, significantly improving student engagement and conversion rates.
Developed on a cloud-native microservices architecture with Node.js, Python, Docker, and Kubernetes, the system ensures scalability, security, and resilience. Data is protected through AES-256 encryption and TLS 1.3, fully compliant with LGPD (Brazilian Data Protection Law).
In production since June 2025, the pilot for second-degree admissions demonstrated significant efficiency gains, reducing analysis time from 7 days to less than 2 hours and lowering staffing needs by 36%. The solution is now being expanded to the EAD transfer process, which represents 50% of all academic analyses at Multivix.
This initiative exemplifies how AI can enhance operational excellence and student experience simultaneously — transforming academic admissions into a fast, transparent, and intelligent journey.
- Project geography
- The project spans the entire Multivix Group, a major higher education network in Brazil with over 60,000 students, 7 on-campus units, and more than 300 distance-learning centers nationwide. The AI-powered admissions system was first deployed in the Second-Degree process and is now being expanded to Distance Learning (EAD) transfers, which account for nearly 50% of all academic evaluations. Its cloud-native architecture enables scalability across all units and remote centers, ensuring a consistent and intelligent digital admissions experience. Recently, Galileo Global Education, one of Europe’s largest education groups, acquired a majority stake in Multivix — reinforcing the project’s international significance and potential for global scalability.
- Additional presentations:
- AI-Powered Academic Admissions- From Documents to Enrollment.pdf