Taleemabad

Organisation Profile

Overview

Taleemabad has been at the forefront of improving learning quality for the past eight years. They reach over 10 million children weekly through digital educational content delivered via public TV & radio, mobile apps, and schools throughout Pakistan.

What type of products are you working on?

We are working on three streams of products, all in early stages:

  1. Automated Exam Checking: Teachers can focus on direct instruction by reducing time spent on administrative tasks like grading assessments and data compilation.
  2. Rapid Educational Video Generation: Creating engaging, pedagogically sound, animated videos that are:

a. contextualised for diverse needs
b. created easily with a single learning outcome-based prompt
c. developed through an easy-to-use AI tool accessible by teachers even in the most underserved communities

3. Automated Textbook/Learning Material Generation: Developing a drag-and-drop textbook creation tool that creates customised textbooks/learning material for each teacher/class, enabling educators and virtually anyone to create AI-powered learning material.

What has surprised you about using AI for education?

Unimaginable speed of innovation: Condensing years of progress into months (and potentially in hours soon) with rapid iteration.

Accessibility of product creation: Creating impactful products is now accessible to anyone with basic knowledge, skills, and a will to make a difference.

Potential to eliminate inequity: Through AI-powered tools, we can empower underserved communities to access highly customised quality education.

Where have you seen the biggest impact for teachers/ students?

The biggest impact we can create is to allow teachers and independent creators to design a variety of educational resources and learning contents, breaking the monopoly of large institutions on quality learning materials. This allows teachers to create an enriched learning environment for their students with materials dynamically tailored to their learning pace and cultural context, leading to improved comprehension and retention.

What are you hoping can be achieved through collaboration with the community?

Shared Expertise and Resources: Collaborating in the open capitalises on a collective pool of knowledge and pre-built tools, speeding up development and avoiding duplication of effort.

Feedback-Driven Development: Collaborative development facilitates a steady stream of user feedback. This ensures that the AI tools being created meet real-world educational needs and preferences.

How can a community help you?

A community can help by providing valuable feedback and insights to improve the AI tools we are developing. They can also contribute shared expertise and resources and collaborate on the development process. Additionally, they can assist in addressing specific challenges, such as grading non-textual content, overcoming limitations in video generation, and advancing multimodal learning material creation.

Moreover, we are eager to seek out new use-cases to better cater to our users. Any feedback, suggestions, or collaborative opportunities in this regard are warmly welcomed.

What are the biggest challenges you are currently facing?

Grading non-textual content (images, diagrams, or pictures) is challenging. We experimented with embedding LlaVa-v1.5-13b (an advanced AI model designed to understand and process both visual content and language) within our current tool to describe images for GPT to later process. However, LlaVa struggles with reasoning.

Creating engaging and pedagogically sound videos automatically is challenging. The process requires precise scripting and VoiceOver synchronisation. Additionally, there are difficulties in training image-generation models to maintain consistency and include text within images.

Combining text, images, and layouts generated by different models into a coherent textbook or learning material remains challenging, requiring advancements in multimodal AI capabilities.

Using generative AI and voice recognition to produce teaching materials and to support assessment.

Profile
Audio: Text2Speech
Large Language Models
Large Multi-Modal Models
NLP: Generation
ML: Classification
MM: Image2Text
MM: Text2Image
MM: Text2Video

Based in Pakistan

U

Developing products for Pakistan

Contact

Usman Javed

i

Strategic Partnerships Lead

Collaborating in the open capitalises on a collective pool of knowledge and pre-built tools, speeding up development and avoiding duplication of effort.

Usman Javed

Strategic Partnerships Lead, Taleemabad

Learning By Doing – Taleemabad

We are providing small grants to several organisations – including Taleemabd – to support the development of AI products & components in LMICs.

We know that innovation investment is high-risk. Our aim is that our community can benefit from the lessons learned in these pilots – what works and what doesn’t.

Learn more about our pilot projects here. We will be following each project and reporting on key learnings.

Sign Up

Join our mailing list to keep up to date with news and events.

Community          Knowledge          FAQ
          Privacy Policy

AI-for-Education.org was set up by Fab Inc. in partnership with Team4Tech. We are grateful to the Bill & Melinda Gates Foundation and the Jacobs Foundation for their support.

Powered by FabData.IO