Zindi Global Artificial Intelligence and Data Analytics Competition 2026

Zindi is running the Financial Inclusion in Africa machine learning competition, a learning challenge to predict which individuals are most likely to have or use a bank account in Kenya, Rwanda, Tanzania and Uganda. It is open to participants in all countries and remains open without a closing date; entrants can compete individually or in a team.

About the Zindi Financial Inclusion Competition

The challenge asks participants to build a machine learning model that predicts whether an individual has or uses a bank account. Its intended purpose is to offer an indication of financial inclusion in the four countries and provide insight into factors connected with individuals’ financial security. The evaluation metric is Mean Absolute Error. In the target labels, 1 indicates that an individual has a bank account and 0 indicates that they do not.

Submissions should follow the specified file structure, including a unique_id and a bank_account column. Participants may make a maximum of 10 submissions per day. A submission counts only after it passes processing; a failed submission is not scored and does not count toward the daily limit. The challenge does not have a public/private leaderboard split.

This is a learning competition, not a prize competition. There are no prizes beyond the knowledge participants gain. Zindi awards 25 points for a participant’s first submission and 50 points for their first non-sample submission. The competition page recommends the challenge to people who are new to machine learning, data science or Zindi, and points to a tutorial on entering a first Zindi competition. Participants interested in related challenges can browse AI opportunities and other competitions and awards.

Entrants may participate alone or form a team of up to four people. When a team is created, its submission allowance is affected by the number of submissions already made by its members: the team receives the competition maximum minus the highest submission count among its members at formation. Multiple accounts per user and collaboration or membership across multiple teams are not permitted. Code must not be shared privately outside a team; code that is shared must be made available to all participants through the platform, such as on the discussion boards.

Solutions must use publicly available, open-source packages and only the datasets provided for the competition. Participants may not use the provided metadata or automated machine learning tools such as AutoML. Pretrained models are permitted if they are openly available to everyone. Data may be stored and used on cloud platforms such as Google Colab or AWS only if it remains private and the use complies with Zindi’s rules. Competition data must not be transmitted, duplicated, published, redistributed or made available to anyone who is not participating. Solutions must not infringe third-party rights. For computer vision challenges, image metadata may not be used; the source states this rule conditionally for challenges of that type.

Who Should Apply

This challenge is suited to people interested in data science, machine learning and predictive modeling, including beginners looking for an introductory competition. It is open regardless of country, and participants can choose to work alone or with a team of up to four. The page does not specify educational, work experience or language requirements.

Key Highlights

• Focus: predicting bank account access or use with machine learning.
• Data context: Kenya, Rwanda, Tanzania and Uganda.
• Deadline: ongoing; the competition does not close.
• Participation: individual or team entry, with teams of up to four.
• Prizes: no prizes; 25 points for the first submission and 50 for the first non-sample submission.
• Submission limit: up to 10 per day.

Participants are subject to the competition rules and Zindi Terms of Use. Zindi reserves the right to update the timeline and rules and to disqualify participants or submissions that it believes violate them.