In team-based courses, instructors often struggle to collect meaningful peer and self-evaluation data while maintaining privacy, reducing administrative overhead, and ensuring students complete evaluations accurately and fairly. Existing tools were either too rigid, lacked instructional flexibility, or required storing sensitive student data.

Traditional learning management systems often struggle to support personalized, reflective, and competency-based learning, especially when formative feedback and learner agency are central goals. To address this gap, I designed and developed an AI-powered Learning Journey platform that guides learners through structured instructional blocks while maintaining flexibility, persistence, and meaningful feedback.

I architected and implemented a modular LMS using JavaScript, RESTful APIs, and local state persistence. The system dynamically delivers six instructional blocks (Intro, Explore, Engage, Practice, Reflect, Assess), integrates generative AI for real-time formative feedback and assessment, tracks learner progress across sessions, and produces a comprehensive learning report. The design emphasizes instructional integrity, reflective learning, and seamless UX, while remaining scalable and framework-agnostic.

The platform successfully demonstrates how generative AI can be embedded responsibly into instructional design to support self-regulated learning, competency-based assessment, and learner reflection. It enables personalized feedback at scale, preserves learning data across sessions, and offers a reusable architecture for educational institutions and training programs seeking intelligent, learner-centered solutions.

I developed a machine learning tool to help educators and instructional designers select appropriate instructional design frameworks based on real-world teaching scenarios. Many practitioners struggle to match specific instructional problems with the most suitable theoretical models, creating a need for an accessible, automated recommendation system. To address this, I created a labeled dataset of instructional cases spanning 11 frameworks and trained multiple models, including TF-IDF + Logistic Regression and Linear SVM, ultimately achieving approximately 93% cross-validation accuracy. I then built a Streamlit application that allows users to input an instructional scenario and receive a recommended framework, supported by explanations and top-three alternatives. This project demonstrates how AI and learning analytics can translate complex educational theory into practical, scalable decision-support tools for teachers, instructional designers, and academic administrators.

During my instructional design internship with Purdue University’s Ti-BOT program, I created a custom-trained GPT assistant to support students working through a competency-based learning system. Students often struggled to interpret competency requirements independently, creating a need for an AI tool that could guide, not answer, while promoting self-regulated learning. To meet this need, I wrote and curated over 360 pages of instructional material in GPT-friendly language, ensuring the assistant delivered scaffolded, inquiry-driven responses aligned with Ti-BOT’s pedagogical goals. The AI assistant is currently being tested with students and has received positive feedback from learners, faculty, and colleagues for its clarity, usability, and strong instructional alignment. This tool is going to be used by +1000 learners.

Link to GPT Assistant:

https://chatgpt.com/g/g-68765c615cf08191a67ad68b21fd196a-tibot-ai 

Link to Documentation:

https://drive.google.com/drive/folders/1QIU8PCd9FMRHCoxHhXCHiOFmYOYzx5Te?usp=drive_link 

I developed a custom Generative AI assistant to support participants during a professional development event hosted by the Virtual World Institute (VWI). The event needed an accessible, real-time support tool that could answer participant questions, recommend resources, and reflect the institute’s focus on immersive, AI-enhanced learning experiences. To address this need, I curated domain-relevant content, trained the assistant, and configured its conversational behavior to align with the event’s learning goals. The AI tool was successfully deployed during the session, enhancing participant engagement and providing on-demand guidance throughout the professional development experience.

Link to the GPT Assistant:

https://chatgpt.com/g/g-6833ef8dade08191a9dca6ca7386ab20-vwi-vanguard-award 

I designed and developed an online asynchronous course to help educators and instructors deepen their understanding of Virtual Reality and learn how to integrate this technology into their teaching. Created as part of my voluntary work with the Virtual World Institute (VWI), the course addresses the need for accessible professional learning opportunities in emerging educational technologies. I built the full learning experience, from instructional design and content development to multimedia production, including an explanatory video I created for the course. Learners progress through modules that introduce core VR concepts, classroom applications, and integration strategies, culminating in a lesson plan submission evaluated by expert reviewers at VWI. Those who meet the performance criteria earn a formal certificate recognizing their competency in VR integration.

I developed a specialized social networking web application designed for educators and instructional designers to connect, share resources, and collaborate professionally. I created the project to address the need for a dedicated online space where instructional design practitioners could exchange ideas beyond general-purpose social platforms. Taking full responsibility for the technical development, I built the application using Python and the Django Framework, implementing user accounts, posting features, and a clean interface optimized for professional networking. The final platform offers a focused, community-driven environment that supports meaningful interaction and knowledge-sharing within the education and instructional design fields.a

I produced an educational video featuring in-depth conversations with two experienced computer security professionals to address the growing need for accessible, research-informed cybersecurity guidance. Many learners and everyday users struggle to protect their self-managed computers, often lacking reliable sources of expert advice. To meet this need, I interviewed the experts, organized the content, and translated their insights into a clear, engaging format that highlights practical strategies for strengthening digital defenses. The final video offers viewers actionable tips, real-world perspectives, and foundational principles to help them navigate the evolving cybersecurity landscape with confidence.

Working with Dr. Marisa Exter, I created an explanatory video for the Deepening Employer, Academia Partnership (DEAP) project, an NSF-funded initiative focused on defining professional competencies in computing. The project needed a clear and accessible explanation of its three-part competency framework, content knowledge, technical and cross-disciplinary skills, and professional dispositions, to help educators and employers develop a shared understanding of workforce readiness expectations. To address this need, I produced a video that translates the project’s research into engaging, real-world examples, such as how a software developer uses both skills and dispositions when troubleshooting code. The final video supports DEAP’s mission by making the framework more understandable and by helping academic programs integrate these competencies into curriculum design through the resources available on the DEAP website.

I developed a Java-based application that enables IELTS candidates to practice the listening test on their own schedule and receive instant, automated feedback. I created the tool to address the need for flexible, accessible listening preparation among my students, many of whom struggled to find reliable practice opportunities outside the classroom. Taking full responsibility for the design and development process, I built the application’s interface, listening modules, and feedback system to closely simulate the IELTS experience. The final product has been used by more than one hundred of my students, who received the app as a complimentary resource to support their exam preparation.

I developed an instructional guide on creating magnetic laminated materials to support hands-on learning activities in university-level Computing Science programs. The idea originated from the creative vision of Dr. Marisa Exter, who recognized the need for sustainable, reusable materials that enhance student engagement. Building on this vision, I designed and produced a comprehensive step-by-step guide that teaches students how to create magnetic laminated components for classroom activities and prototyping tasks.

By integrating practical craftsmanship with environmentally conscious practices, the guide encourages students to adopt reusable materials and reduce waste. The final product not only enriches hands-on experiences within the Computing Science discipline but also promotes sustainable thinking and resource-efficient design.

designed and developed the official website for the Virtual World Institute (VWI) as part of my voluntary role as the Director of IT and Technology. In addition, I implemented Moodle as the institute’s Learning Management System (LMS), enabling VWI to deliver its courses and training programs online.

Website: www.virtualworldinstitute.org

Tavakoli. J. Garage Door Opener / 2022 

Iran, Patent Application No: 140050140003007883. Patent and Trademark Organization

Cars commonly store garage door remotes, creating security risks and relying on disposable batteries. I set out to design an integrated, energy-efficient alternative.

I developed an electronic board that connects directly to the vehicle’s power system and repurposes an unused dashboard button to activate the garage door. I led the entire invention process, from ideation and system architecture to schematic development and PCB design, using Altium Designer to create the circuit board.

The final solution eliminates external garage remotes, increases vehicle security by removing a common theft target, and reduces environmental waste by drawing power from the car’s existing electrical system.