ISMAIL ISMAIL TIJJANI

Mechatronics Engineering Graduate | Applied Computer Vision Researcher | MLOps Practitioner

About Me

I am a Mechatronics Engineering researcher and developer focused on computer vision systems designed for real-world deployment. My work is guided by a data-centric and deployment-aware approach to building intelligent systems that operate reliably in resource-constrained and underrepresented environments.

My technical experience spans the full pipeline of intelligent system development, from dataset collection and annotation to deep learning model design, including object detection, instance segmentation, and OCR systems, as well as elements of MLOps for model training and deployment. I am particularly interested in the integration of perception systems with practical deployment constraints, with applications in agriculture, medical imaging, and intelligent infrastructure monitoring.

I am also the founder of EJAZTECH.AI, an initiative focused on bridging the gap between local AI landscape and the global ecosystem.

Research Interests

My research focuses on developing efficient and robust mechatronic systems integrated with artificial intelligence for real-world, resource-constrained environments. I am particularly interested in data-efficient computer vision, where models are designed to perform reliably under limited computation and small or imperfect datasets.

I am also interested in embodied AI and intelligent robotic systems, with an emphasis on integrating perception models into physical agents. In addition, I am interested in multimodal learning approaches that improve adaptability in complex, real-world scenarios.

Education

Bayero University Kano Bachelor of Engineering in Mechatronics 2020- 2026

Publications

NLPDRS: A YOLOv8 and PaddleOCR-Based End-to-End Framework for Nigerian License Plate Detection and Recognition System

II Tijjani, SM Ibrahim, AA Mustapha (Oct, 2025)

https://ijsrcse.isroset.org/index.php/j/article/view/748

This localized, end-to-end framework tackles the failure of standard vehicle tracking systems against the non-uniform fonts and unpredictable lighting typical of Nigerian license plates. By building a custom-annotated dataset and leveraging YOLOv8 for real-time plate localization, the system utilizes an optimized OCR pipeline enhanced by k-means clustering and adaptive thresholding to deliver highly accurate alphanumeric extraction. The resulting model achieves a 0.93 mAP@0.5, 0.91 recall, and 0.87 precision, providing a scalable solution ready for real-world deployment in local traffic monitoring, automated access control, and national security infrastructure.

Performance Comparison of Deep Learning Techniques in Naira Classification

II Tijjani, AA Mustapha, IT Idris (Oct, 2024)

https://arxiv.org/abs/2412.02072

This study addresses the challenge of automated currency recognition by deploying and evaluating deep learning architectures to classify Nigerian Naira notes by denomination. Utilizing a diverse custom dataset of 1,808 images captured under varying environmental conditions, the framework evaluates multiple models, with MobileNetV2 achieving peak performance, yielding a 90.75% training accuracy and an 87.04% validation accuracy. Optimized for efficiency and light footprint, this model provides a scalable, edge-ready solution designed for real-world integration into automated cash sorting systems, vending hardware, and assistive technologies for the visually impaired, ultimately enhancing financial transaction security and efficiency.

Conference Papers

Data-Driven Approach in Reservoir Production Forecast: Comparison of Machine Learning and Deep Learning

SI Muhammad, N Makwashi, A Abdulsalam, II Tijjani (Aug, 2025)

Presented at the SPE Nigeria Annual International Conference and Exhibition

https://doi.org/10.2118/228783-MS

This data-driven study tackles the economic challenges of declining hydrocarbons and rising unwanted fluid production by optimizing production forecasting through machine learning and deep learning. By evaluating regression models (Decision Tree and Random Forest) against sequential time-series models (RNN and LSTM), the framework leverages key operational inputs, such as choke size and various pressure metrics, to forecast oil, gas, and water outputs simultaneously. All architectures achieved robust performance with R² scores exceeding 0.85, while the LSTM model proved superior with a peak R² of 0.95 due to its strength in capturing long-term temporal dependencies in reservoir data. The resulting framework provides a highly accurate, computationally efficient alternative to traditional reservoir simulation, offering a scalable asset for strategic production planning and cost-effective reservoir management.

Evaluating Vision-Language Models as a Zero-Shot Learning Alternative to You Only Look Once and Optical Character Recognition for Nigerian License Plate Recognition

II Tijjani, AA Mustapha, SI Muhammad, MB Aliyu (Nov, 2025)

Presented at ICCAIT (International Conference on Computing and Artificial Intelligence Technologies)

Presented by: Self

This study evaluates the potential of Vision-Language Models (VLMs) as a unified, zero-shot alternative to traditional, resource-heavy multi-stage YOLO and OCR pipelines for license plate recognition. Utilizing a curated benchmark dataset of 88 challenging real-world images from Nigeria, the research conducts a comparative analysis across five leading architectures: Gemini 2.0 Flash, Qwen2.5-VL-7B-Instruct, GPT-4o, Claude 4 Sonnet, and Llama 3.2 Vision 90b. Evaluated via Character Error Rate (CER), the findings reveal that Gemini and Qwen significantly outperform competing models in both accuracy and robustness under unstructured environmental conditions. The work provides critical insights into the practical advantages of zero-shot VLMs in regional surveillance, while objectively challenging provider performance claims in complex, real-world edge scenarios.

Sesame Plant Segmentation Dataset: A YOLO-Formatted Annotated Dataset

SI Muhammad, II Tijjani, SY Jumare, FI Jibrin (Nov, 2025)

Presented at ICCAIT (International Conference on Computing and Artificial Intelligence Technologies)

Presented by: Co-author

This paper introduces an open-source, pixel-level annotated dataset designed to advance computer vision applications for sesame cultivation during crucial early growth stages. Compiled from high-resolution imagery captured under diverse environmental conditions across farms in Katsina State, Nigeria, the dataset features images structured in a YOLO-compatible format and annotated via the Segment Anything Model (SAM 2) under farmer supervision. Evaluated using the Ultralytics YOLOv8 framework, the baseline models achieved strong results across both tasks, yielding an 84% detection mAP@0.5 and an 84% segmentation mAP@0.5. By shifting from standard bounding boxes to precise instance segmentation, this localized resource provides a foundational framework for real-world agricultural automation, including autonomous weeding, plant health monitoring, and precise yield estimation.

Selected Projects

Autonomous AI-Powered Three-Wheeled Herbicide Sprayer

Designed and manufactured a customized three-wheeled robotic platform optimized for stability during localized agricultural operations. The system deploys a camera to feed real-time visual frames into an edge-optimized YOLO instance segmentation framework that detects weeds during crucial early plant growth stages. The mechatronic control pipeline processes these coordinates instantly, executing localized micro-interventions by triggering target-specific spray actuators to eliminate the weed while conserving chemical volume and protecting adjacent crops.

Nigeria License Plate Detection and Recognition System (NLPDRS)

A robust vehicle identification framework featuring a custom YOLOv8 model for real-time license plate localization and an optimized character recognition engine to extract non-uniform text. Engineered to handle regional plate variations and complex lighting, the system integrates an intelligent mechatronic hardware layer utilizing IR sensors for vehicle detection and dynamic database querying for instant plate validation. Upon verification, the system updates security records and provides real-time feedback, triggering an onboard buzzer alert if an unauthorized or flagged vehicle is detected.

GitHub Page: https://github.com/esssyjr/Nigeria-License-Plate-Detection-and-Recognition-System-NLPDRS

Video Demo: https://www.linkedin.com/feed/update/urn:li:activity:7227051054800990210

MPOX Detection API

A specialized deep learning framework and inference API optimized to accelerate clinical workflows by detecting MPox lesions from dermatological and medical imagery. Built to provide reliable diagnostic support, this system can be integrated into telemedicine platforms and mobile health applications to assist frontline workers with rapid triage.

Github Page: https://github.com/esssyjr/MPOX_DETECTION

AI-Powered Naira Classification Model

A deep learning classification model trained on a diverse dataset of Nigerian banknotes for robust recognition under non-ideal conditions. Optimized for edge deployment, the system integrates an end-to-end MLOps pipeline using MLflow and Airflow for experiment tracking, automated retraining, and continuous deployment. It also includes a web interface and a human-in-the-loop validation system for feedback-driven model improvement.

GitHub Page: https://github.com/esssyjr/NAIRA_CLASSIFICATION

Video Demo: https://www.linkedin.com/feed/update/urn:li:activity:7207685557924147200/?originTrackingId=sjuhZOhuTW%2BjDscnAWQM7Q%3D%3D

AgriAI: Weed Detection & Crop Health Monitor

An agricultural vision platform that addresses crop management challenges by combining real-time weed detection and health monitoring into a unified system. Developed using precise localized field data, the solution provides an accessible interactive web dashboard designed to assist farmers with target-specific interventions and automated field management.

GitHub Page: https://github.com/esssyjr/Plant_Detection_and_their_Health_Conditions

Website Demo: https://huggingface.co/spaces/esssyjr/Plant_Detection_and_Their_Health_Condition_FCMB_HACKATHON

Video Demo: https://drive.google.com/file/d/14aV22Wq9OXHIeZM6AwJSG6rfe1_9zJJq/view?usp=drive_link

Work Experience

Founder and President, EJAZTECH.AI

July 2023 – Present

Founded and led an AI initiative focused on developing locally relevant datasets and AI solutions.

Mentored 3,000+ students in AI through structured training programs and workshops.

Led development of multiple real-world AI systems across agriculture, security, and healthcare.

Coordinated collaborative research and engineering teams for AI solution development.

Computer Vision Developer, InnovateHealth Africa

September 2024 – Present

Designed and deployed computer vision models for healthcare applications, including disease detection and analysis systems.

Led dataset curation and annotation pipelines for high-quality medical imaging datasets.

Collaborated with cross-functional teams to integrate AI solutions into real-world healthcare workflows.

Contributed to scalable AI system design for practical deployment.

Research Assistant (Computer Vision), Germanium Solution

December 2024 – May 2025

Assisted in the core research and development of an AI-based malaria detection framework, contributing across the entire scientific pipeline from raw microscopic data collection to preprocessing and feature extraction.

Led rigorous medical data curation and annotation workflows to construct high-quality datasets for training diagnostic deep learning models.

Conducted model experimentation and evaluation for medical image analysis to advance early-stage pathogen detection and automated diagnosis.

Deployed computer vision architectures into CCTV-based monitoring systems for real-time intelligent surveillance, crowd metrics, and edge anomaly detection.

Dataset Contributions

Nigerian Currency (NAIRA) Classification Dataset

A curated image dataset of Nigerian Naira banknotes captured under real-world conditions using mobile devices. The dataset includes seven denominations (₦5–₦500) and is structured into training, validation, and testing splits for robust evaluation of image classification models.

This dataset supports research in currency recognition systems for automated financial processing, assistive technologies, and edge AI deployment in low-resource environments.

https://www.kaggle.com/datasets/ismailismailtijjani/naira-nigerian-currency-dataset

Nigerian Tricycle (KEKE NAPEP) Detection Dataset

A YOLO-formatted object detection dataset containing images of commercial tricycles collected across urban environments in Kano State, Nigeria. The dataset includes diverse real-world conditions such as traffic congestion, varying lighting, and multiple viewpoints.

It is designed for object detection and tracking tasks in intelligent transportation systems, urban mobility analytics, and traffic monitoring applications.

https://www.kaggle.com/datasets/ismailismailtijjani/keke-napep-tricycle-dataset

Sesame Plant Segmentation Dataset

A pixel-level annotated agricultural dataset focused on sesame plant segmentation in real farm environments. The dataset captures multiple growth stages, lighting conditions, and field perspectives, enabling precise instance segmentation for agricultural AI systems.

It supports applications in crop monitoring, yield estimation, and automated agricultural intervention systems.

https://www.kaggle.com/datasets/ismailismailtijjani/sesame-plant-detection-dataset

AgriAISeg Multicrop Segmentation Dataset

A large-scale agricultural image dataset designed for pixel-level segmentation of multiple crop types, including tomato, cabbage, and sesame. The dataset is structured for deep learning-based segmentation tasks and supports scalable training for agricultural vision systems.

It is intended for research in precision agriculture, automated farming systems, and plant-level monitoring.

https://www.kaggle.com/datasets/ismailismailtijjani/agriaiseg

Oil & Gas Production Monitoring Dataset

A structured tabular dataset containing surface and subsurface production parameters from oil wells. It includes variables such as pressure, temperature, flow characteristics, and production outputs (Water Cut and BS&W).

This dataset supports machine learning applications in production forecasting, reservoir analysis, and energy systems optimization.

https://www.kaggle.com/datasets/sgobir/production-dataset

Projects

Food Vision

An end-to-end multimodal computer vision application that detects Nigerian dishes in real time using Roboflow and integrates Google’s Gemini AI to generate customized nutritional insights and health benefits. Deployed via a Hugging Face space, the interactive Gradio interface includes a text-to-speech module to read dietary recommendations aloud for enhanced user accessibility.

Video Demo: https://youtu.be/WsCxMQOw78I?si=5SDpxRre045zTSj-

Website Demo: https://huggingface.co/spaces/esssyjr/FOOD

GitHub Page: https://github.com/esssyjr/Food-Vision

Halal View

An AI-powered automated video content moderation pipeline engineered to identify specified visual contexts in media streams and apply precise, dynamic Gaussian blurring. Designed to safeguard viewer preferences and maintain cultural compliance, this system provides a scalable solution for television networks, digital broadcasting, and parental control applications.

Video Demo: https://youtu.be/_1HlX8Evazo?si=TELF7BqYAT9p9k9j

GitHub Page: https://github.com/esssyjr/MPOX_DETECTION

Keke Napep Detection and Tracking System

A high-throughput computer vision pipeline engineered to detect, segment, and track commercial tricycles (Keke Napep) across real-world urban video feeds. Tailored to address the specific layout of Nigerian transit ecosystems, the system outputs precise localization and count data to optimize municipal fleet management, toll automation, and traffic flow analytics.

GitHub Page: https://github.com/esssyjr/KEKE_NAPEP_DETECTION_AND_TRACKING_SYSTEM

Video Demo 1: https://youtu.be/sZ4QVAU8XIg?si=vBMNiVAFNJgGR16n

Video Demo 2: https://youtu.be/UOqrPQgLDT8?si=hdb-toFdPTlIlgwO

Website Demo: https://huggingface.co/spaces/esssyjr/KEKE_NAPEP_DETECTION

Sports Image Classification System

An image recognition framework utilizing MobileNetV2 transfer learning to perform highly efficient classification across one hundred distinct sports categories. By exploiting pre-trained convolutional feature maps, the model minimizes computational overhead while maintaining high accuracy, making it ideal for mobile apps and low-power devices.

GitHub Page: https://github.com/esssyjr/SPORTS_CLASSIFICATION

Surface Crack Classifier

A structural health monitoring model designed to classify infrastructural surfaces into binary states of integrity based on the presence of structural fractures. Engineered to automate routine physical inspections, this model helps flag critical defects early across concrete roads, facility walls, and public bridge structures.

GitHub Page: https://github.com/esssyjr/SURFACE_CRACK_CLASSIFICATIONS

6-Places Scene Classifier

A deep neural network trained to execute environmental scene categorization, identifying images across distinct topographical classes including buildings, forests, glaciers, mountains, seas, and streets. The system provides foundational spatial context layers useful for geoguessr engines, map organization tools, and autonomous vehicle navigation.

GitHub Page: https://github.com/esssyjr/6-PLACES-CLASSIFICATION

4-Animals Classifier

A lightweight convolutional neural network built as a comparative exploration of model scaling, designed to classify image vectors of cats, dogs, elephants, and giraffes. The codebase showcases optimization techniques, pipeline configurations, and data augmentation practices used to stabilize training accuracy with localized datasets.

GitHub Page: https://github.com/esssyjr/4-Animals-Classification

Predictive Model for Consumer Behavior

A machine learning exploration modeling consumer purchasing actions through a comparative analysis of Decision Trees, Logistic Regression, and simple feedforward Artificial Neural Networks. The repository outlines complete structured data preprocessing workflows, multi-feature correlation analysis, and comparative classification evaluation metrics.

GitHub Page: https://github.com/esssyjr/Binary-Classification-Bike-Buyer

Multi-Class Dry Bean Classifier

A dense feedforward neural network optimized to categorize agricultural varieties of dry beans based on geometric, morphological, and structural features extracted from seed imagery. The pipeline focuses heavily on feature scaling and engineering to resolve high-overlap classification challenges across related biological strains.

GitHub Page: https://github.com/esssyjr/Multi-class-Classification-of-Dry-Beanss

DCARS: Automated Diagnostic Center Reception

An intelligent, production-oriented automation system designed to streamline patient intake and data flows within medical diagnostic facilities. The software provides a structured solution to optimize patient records management, queue scheduling, and internal front-desk coordination to eliminate operational administrative bottlenecks.

GitHub Page: https://github.com/esssyjr/DCARS

Articles

Day 4: Adding Human-in-the-Loop Annotation to the MLOps Pipeline

Concludes the MLOps pipeline series by establishing a Human-in-the-Loop (HITL) microservice to catch edge-case misclassifications and eliminate silent model degradation. It outlines an asynchronous feedback loop using shared Docker volumes where human-corrected data is systematically structured and automatically fed back into Airflow retraining cycles.

https://medium.com/@ismailismailtj/day-4-adding-human-in-the-loop-annotation-to-the-mlops-pipeline-6da85777049e

Day 3: Deploying Containerized ML Services with Docker, FastAPI, Streamlit, Airflow & MLflow

Demonstrates how to containerize and orchestrate a multi-service production pipeline using Docker and Docker Compose to eliminate environment mismatches. It breaks down thread-safe, lazy model loading in FastAPI via MLflow registry aliases alongside an interactive Streamlit frontend to achieve seamless, zero-downtime model updates.

https://medium.com/@ismailismailtj/day-3-deploying-containerized-ml-services-with-docker-fastapi-streamlit-airflow-mlflow-6584b9b4d98b

Day 2: Orchestrating Production ML Model Retraining with Apache Airflow

Focuses on mitigating production model degradation due to data drift by automating daily retraining pipelines with Apache Airflow. The piece details DAG architecture, PythonOperator task dependencies, and robust error-handling mechanisms designed to keep production models accurate without manual intervention.

https://medium.com/@ismailismailtj/day-2-orchestrating-production-ml-model-retraining-with-apache-airflow-06de179a0913

Day 1: Model Experimentation with MLflow

The first installment of an end-to-end MLOps series detailing the technical implementation of MLflow to eliminate ad-hoc, unorganized machine learning workflows. It provides an architectural walkthrough of centralized tracking servers, backend metadata storage, and automated TensorFlow logging to ensure complete experiment reproducibility.

https://medium.com/@ismailismailtj/day-1-model-experimentation-with-mlflow-711584f28195

Day 0: Building a Real End-to-End MLOps Pipeline

Serves as the foundational primer for an incremental MLOps series, addressing the industry-wide challenge of machine learning models failing to transition from static Jupyter notebooks to active production environments. It introduces the architectural blueprint and operational roadmap for a reliable, living system powered by an industry-grade technology stack. Utilizing a real-world Nigerian currency dataset, the article outlines the design requirements for automated data ingestion, human-in-the-loop validation, automated retraining, and zero-downtime deployment.

https://medium.com/@ismailismailtj/day-0-building-a-real-end-to-end-mlops-pipeline-9c6aa5ed4df6

AI Education: The True Engine of Development

Explores the critical role of technical mentorship and foundational AI education in shifting users from passive consumers to proactive creators. Based on outreach experiences with EJAZTECH.AI, it advocates for demystifying technology to transform industry fear into practical, localized innovation.

https://medium.com/@ismailismailtj/ai-education-the-true-engine-of-development-627ad2698ea3

AI Needs a Body — And Mechatronics Provides It

Argues that the next frontier of artificial intelligence will unfold in the physical world where software must interact with reality. It highlights Mechatronics Engineering as the essential hardware foundation—providing the sensors, actuators, and control systems that serve as the physical body for AI brains to see, predict, and act.

https://medium.com/@ismailismailtj/ai-needs-a-body-and-mechatronics-provides-it-c606b6d191a3

Integrating Gemini, Roboflow, Gradio, and Hugging Face to Build AI-Powered Systems

A technical guide demonstrating how to build a localized, multimodal "Food Vision" system. The article details end-to-end integration of Roboflow for object detection, Google’s Gemini for tailored nutritional insights, and Hugging Face and Gradio for deploying an interactive web interface complete with audio output.

https://medium.com/@ismailismailtj/integrating-gemini-roboflow-gradio-and-hugging-face-to-build-ai-powered-systems-887f4d91f60b

Building Africa’s AI Future Starts with Data Collection: A Call to Action

An analysis of why the African continent risks lagging behind in the global AI revolution due to a critical deficit in localized, digitized data collection. It argues against the ineffective copy-pasting of Western AI models, emphasizing that impactful solutions must be built from the ground up using infrastructure-specific, regional data.

https://medium.com/@ismailismailtj/building-africas-ai-future-starts-with-data-collection-a-call-to-action-645d2438aee5

AI Beyond Coding: A Data-Centric Paradigm

A deep dive into why code accounts for only about 30 percent of the machine learning pipeline, shifting the focus toward Andrew Ng’s data-centric AI philosophy. The piece breaks down project scoping, rigorous data preprocessing, model deployment, and MLOps, proving that high-quality, diverse datasets outperform minor architecture tweaking.

https://medium.com/@ismailismailtj/ai-beyond-coding-a-data-centric-paradigm-7f4577748c5a

Path to Artificial General Intelligence (AGI) 2

A conceptual critique of the term "Artificial General Intelligence," advocating for the more precise framing of "Human-Level Intelligence." The article outlines the core cognitive architectures required for true machine autonomy—specifically learning, memory, reasoning, and planning—while emphasizing a gradual, collaborative framework.

https://medium.com/@ismailismailtj/path-to-artificial-general-intelligence-agi-2-6c5051af0baf

How Students Will Use ChatGPT Effectively

A practical guide for students navigating generative AI tools in academia. It explores how to leverage large language models for rapid summarization, research synthesis, and ideation, while strictly outlining strategies for rigorous fact-checking, bias mitigation, and maintaining academic integrity.

https://medium.com/@ismailismailtj/how-students-will-use-chatgpt-effectively-1d7f57d7e722

Understanding How ChatGPT Works and Addressing Raised Concerns

A technical breakdown of transformer architectures and language model mechanics simplified for a broader audience. The article demystifies how probabilistic text generation works while addressing common societal anxieties surrounding AI hallucination, training data ethics, and privacy.

https://medium.com/@ismailismailtj/understanding-how-chatgpt-works-and-addressing-raised-concerns-09f7835d17c0

Path to Artificial General Intelligence (AGI 1)

The foundational installment examining the milestone markers, historical context, and technical evolutionary steps required for machine learning systems to transition from narrow tasks toward generalized problem-solving capabilities.

https://medium.com/@ismailismailtj/path-to-artificial-general-intelligence-agi-1-c07208b13020

AI and the Future of Jobs

An economic and technical analysis of automation’s impact on the global workforce. The piece separates media hype from reality, discussing job displacement, the emergence of the prompt engineering and data curation economy, and the necessity of rapid upskilling.

https://medium.com/@ismailismailtj/ai-and-future-of-jobs-a831e9bdc494

AI is Here for Good

An optimistic, evidence-based perspective on the positive societal impacts of artificial intelligence, highlighting its capacity to democratize healthcare, optimize agricultural yields, and streamline civic infrastructure when developed ethically.

https://medium.com/@ismailismailtj/ai-is-here-for-good-bcdb47d6cebe

Technical skills

Computer Vision

Object Detection, Instance Segmentation, Optical Character Recognition (OCR), Image Classification, Video Analysis

Machine Learning & AI

Deep Learning, Data-Efficient Learning, Transfer Learning, Model Evaluation, Multimodal Learning

Intelligent & Embedded Systems

Edge AI Deployment, Embedded AI Systems, Real-Time Inference Pipelines, Camera-Based Vision Systems, Sensor Integration, Microcontroller-Based Control Systems

MLOps & Deployment

Model Deployment, API Development, Docker, FastAPI, Experiment Tracking (MLflow, Airflow)

Programming & Tools

Python, PyTorch, TensorFlow, OpenCV, Git, Linux

Leadership & Impact

AI Mentorship Program — From Zero to Hero

Founded and led a structured AI mentorship initiative focused on guiding beginners from foundational concepts to practical AI development. The program is currently running its fourth cohort and has supported students through hands-on learning, project development, and technical mentorship.

AI Awareness Outreach

Organize outreach sessions for secondary school students focused on responsible, ethical, and productive use of artificial intelligence technologies.

Monthly AI Awareness Webinars

Host monthly public webinars designed to help non-technical audiences understand practical applications of AI and emerging technological trends.

Python & Machine Learning Workshops

Conduct practical workshops introducing programming, machine learning, and AI fundamentals to young learners, particularly individuals between the ages of 16–20.

Teacher Capacity Building Program

Train educators on integrating AI tools into teaching and learning environments while promoting responsible and ethical AI usage in classrooms.

Awards and Recognition