ELIGIBILITY:

The Challenge Task:

Create or improve a real-world voice AI application

The transcript received can be used to perform a downstream task. Examples include a voice assistant that completes a booking, a voice-driven triage tool, a spoken-language form-filler, or a voice interface that takes an action on a user’s behalf.

Each submission must:

Use voice to achieve a downstream task (the agentic component), not simply convert speech to text.

Benchmark performance on code-switched audio input across at least three speech models, including an Intron Sahara API and at least two other models of the team’s choice (global, local, open-source, commercial, pretrained, or fine-tuned).

Target a specific vertical or use case (see categories below).

Where possible and permitted, submit code-switched audio samples used for testing with basic metadata (language pair, domain, accent/country, device type, and noise conditions).

Categories: Five tracks designed for real-world impact across Africa – Health, Legal & Public Services, Fintech, Telco & Customer Experience, Agriculture & Education, and Other High-Impact Use Cases

Applications are open to builders across Africa and the diaspora. Apply as a solo builder or as a team. We especially encourage:

Developers & engineers

AI/ML researchers

Startups & product teams

Student teams

Open-source builders

NGOs & civic-tech teams

Companies with voice or customer-facing workflows

Bonus points will be given to solutions addressing a real problem for a meaningful population size.
Submissions can include fresh ideas, production apps, apps close to deployment, etc., clearly designed for real-world use.

Who should enter

Applications are open to builders across Africa and the diaspora. Apply as a solo builder or as a team. We especially encourage:

Team Formation:

You can apply as an individual or as a team.

Teams may include developers, researchers, designers, domain experts, product managers, linguists, clinicians, customer-support operators, field workers, or anyone who understands the problem you are solving.

Recommended team size: 2–5 people.

What to Submit:

Keep it concise. Focus on impact, execution, and responsible use.

Problem & Solution: A short description of the problem and your proposed solution.

Working Prototype: A short video demo of a working prototype or integrated application.

Code & Docs: Code or technical documentation for your implementation.

Benchmark Results: Comparing Sahara against at least three other speech models.

Responsible AI: A note on privacy, consent, safety, and responsible data use.

BENEFIT:

$10,000Total Cash Prize Pool

Cash prizes will be awarded to top-performing solutions across the challenge categories.

Workshop Challenge Prize

$1,200: Winner

$800: Runner-up