We are looking for an AI Engineer to help Zeraki use AI in how we build and in what we ship. Reporting to the CPTO and working day to day with engineering leads, Product Managers, Data, Support and Operations, you will start by building AI tools that improve our internal workflows and save engineering time, then help bring AI into the products used by teachers, school leaders, parents and learners across Africa.

This is a hands-on role: you will write code, ship it and learn from how people use it. Zeraki's products share data on learning, school operations and finance in a way few platforms can match, and you will help turn that data into tools and features people rely on. You will get regular code reviews, clear priorities and support from experienced engineers and we expect you to grow quickly in skill and ownership.

Why This Role, Why Now

Lots of data and routine work very little AI in it yet.

Zeraki holds something few platforms in Africa can match: learning, school operations and finance data for thousands of schools. Today very little of it is powered by AI, and much of our engineering time still goes into repeatable work that AI can take on. We need a hands-on engineer who is eager to build with AI, keeps costs low and helps free our teams from repetitive tasks so they can focus on the work that matters most.

QUESTIONS YOU WILL HELP ANSWER

Where is repetitive work costing our engineers and operations teams the most time, and what should AI take on first?

Where can AI make Zeraki genuinely more useful for teachers, school leaders and parents?

How do we turn connected school data into predictions and insight people trust enough to act on?

How do we know an AI system is right, wrong or getting worse before our users do?

How do we protect student and school data, especially data about minors, as AI gets access to more of it?

When should a person review what AI suggests, and when is it safe to automate fully?

What are the costs of running a particular model and how can we bring it down?

What You Will Do

The weights below show where your time and impact are expected to go in your first year the same weights your performance is reviewed against.

01 Internal Workflows · 20% of role

Good looks like: operations, finance, support and sales teams spend less time on routine work because of tools you built.

• Build secure integrations and simple agents that connect AI to Zeraki's internal systems.
• Automate repetitive work for operations, finance, support and sales teams.
• Measure the time your tools save, and improve them based on feedback.

02 Engineering Productivity · 20% of role

Good looks like: engineers spend less time on routine work because of tools you built.

• Build and improve AI tools for triage, debugging, code review, testing and documentation.
• Help engineers across Zeraki adopt the AI tools you build.
• Measure the engineering time your tools save, and improve them based on feedback.

03 AI in Our Products · 20% of role

Good looks like: AI features you helped build are used by teachers, school leaders and parents.

• Build and test AI features with Product, Design and engineering leads.
• Take features from prototype to production, with review and support.
• Build for our context: mobile-first, low-bandwidth, multilingual and multi-curriculum.

04 Data, Prediction & Insight · 15% of role

Good looks like: clean data and clear analysis that help teams make better decisions.

• Build data pipelines and analysis on Zeraki's data.
• Help build and test predictive and forecasting models with Product and Engineering.

05 Quality, Safety & Cost · 15% of role

Good looks like: what you ship is tested, monitored and never puts student or school data at risk.

• Write tests and evals for every AI feature you build, and monitor them in production.
• Build in human review where people check and approve AI suggestions.
• Follow strict data access and privacy rules in line with Kenya's Data Protection Act, with extra care for data about minors.
• Keep model cost and latency in check.

06 Learning & Collaboration · 10% of role

Good looks like: you grow fast, share what you learn and are easy to work with.

• Learn our codebase, systems and how schools use Zeraki.
• Share what you build and learn with engineers across Zeraki.
• Ask for feedback and act on it.

Your scope grows with you: as your skills grow, you will own larger AI systems end to end.

Performance principle

Demos and prototypes are diagnostics. You will be judged on what reaches production, the time and outcomes it delivers for Zeraki and schools, and how safely it runs.

What Great Looks Like

What great looks like in each area of the role.

Area

What great looks like

Internal Workflows

Finds the most time-consuming manual work · ships tools teams actually use · measures time saved

Engineering Productivity

Ships tools engineers actually use · measures time saved · improves from feedback

AI in Our Products

Builds features that solve real user problems · gets them to production · designs for mobile-first, low-bandwidth, multilingual use

Data, Prediction & Insight

Reliable pipelines · clear analysis · models tested before anyone relies on them

Quality, Safety & Cost

Tests and evals on everything shipped · human review where it matters · careful with data · aware of cost

Learning & Collaboration

Learns fast · asks good questions · shares what works · takes feedback well

What Success Looks Like

First 90 days

• Time with Engineering, Product, Support and Operations to learn our platform, data and codebase
• First AI tool shipped internally, with time saved measured
• Working confidently within our engineering, testing and data-protection standards

First six months

• Several AI tools in daily use across engineering or operations
• Contributed to an AI feature live in a Zeraki product
• Tests, evals and monitoring in place for everything you have shipped

First 12 months

• Owning AI features end to end, from idea to production
• A clear time-saved story from the tools you built
• Ready to take on larger AI systems and help others use AI well

Your Scorecard

You are reviewed on the six areas of the role, weighted as below, together with the role objectives and our eight values.

Area

Weight

Internal Workflows

20%

Engineering Productivity

15%

AI in Our Products

25%

Data, Prediction & Insight

15%

Quality, Safety & Cost

15%

Learning & Collaboration

10%

Total

100%

Role objective

Target

AI features you ship with tests and evals in place

100%

AI output reaching users with human review where required

100%

Time saved through the AI tools you build

Measured and improving every quarter

Usage of the AI tools you build

Improving every quarter

Data protection or security incidents caused by your work

Zero

Growth goals reviewed with your manager

Every quarter

What You Bring

Experience & technical skills

• 1–2 years of software engineering experience, including internships.
• A degree in Computer Science or a related field, or equivalent practical experience.
• Solid programming in Python or Java.
• Built at least one project using an LLM API — at work, in study or on your own.
• Familiar with prompting, RAG or tool use, or keen to learn them fast.
• Good grasp of SQL, REST APIs and Git.
• Writes tests and understands why they matter.
• Comfortable cleaning, joining and analysing data.
• Exposure to cloud services, CI/CD, machine learning or MCP (a plus).
• At least one use case you have built using ai agents.

What we are looking for

• A builder — you learn by making things and shipping them.
• Curious and quick to learn — you pick up new tools and ideas fast.
• An owner — you see a manual process and want to remove it.
• AI-fluent — you use AI coding tools daily and can explain how they help you.
• Open to feedback — you ask questions, take reviews well and improve.
• Pragmatic — you care about what works, not what's trendy.
• A clear communicator — with engineers and non-technical teams alike.

Where you might be coming from

You may be a recent Computer Science or engineering graduate, a junior software engineer with a year or two of experience, or someone who has built AI projects through internships, hackathons, Zindi or Kaggle competitions, or open-source work. Your title matters less than what you have built and how quickly you learn.

What This Role Is Not

• Not a research role — You build AI that ships and gets used, not models for papers.
• Not a demo factory — Prototypes matter only when they reach production with tests and monitoring.
• Not a solo role — Your work is reviewed and supported by engineering leads, manager and the CPTO.
• Not static — As AI and Zeraki change, the tools and how we use them will change too.

What We Offer

At Zeraki, you'll have the opportunity to work at the intersection of education, technology and financial services while helping shape one of the company's most strategic product areas.

You'll work closely with executive leadership, experienced product leaders, designers, engineers and leading financial institutions, gaining hands-on experience building products that improve education for millions of families across Africa.

You'll be trusted with meaningful ownership, supported by a collaborative team and challenged to grow as a Product Manager by solving real customer problems and making real product decisions.

Our Culture

At Zeraki, our culture defines how we work, how we make decisions and how we succeed together. We are united by eight core values: Integrity, Love of Learning, Highest Standards, Deliver Results, Hire & Develop the Best, Manage Through Trust, Empathy, and Have Backbone; Disagree & Commit.

We act with integrity, doing what is right and taking ownership of our actions. We have a love of learning, staying curious and continuously seeking to improve. We hold ourselves to the highest standards, continually raising the bar in the quality of our work, and we deliver results by taking ownership, acting with urgency and following through on our commitments.

We hire and develop the best, surrounding ourselves with exceptional people and creating opportunities for them to grow and excel. We manage through trust, giving people the context, responsibility and support to do their best work.

We lead with empathy, creating an environment where people are respected, supported and able to thrive. We have backbone, disagree and commit—we speak up when we disagree, challenge ideas respectfully and, once a decision is made, commit fully to making it successful.

Together, these values create a culture of freedom and responsibility, high performance, continuous learning and meaningful impact, empowering every Zerakian to Be Extraordinary. Copy

• Build secure integrations and simple agents that connect AI to Zeraki's internal systems.
• Automate repetitive work for operations, finance, support and sales teams.
• Measure the time your tools save, and improve them based on feedback.
• Build and improve AI tools for triage, debugging, code review, testing and documentation.
• Help engineers across Zeraki adopt the AI tools you build.
• Measure the engineering time your tools save, and improve them based on feedback.
• Build and test AI features with Product, Design and engineering leads.
• Take features from prototype to production, with review and support.
• Build for our context: mobile-first, low-bandwidth, multilingual and multi-curriculum.
• Build data pipelines and analysis on Zeraki's data.
• Help build and test predictive and forecasting models with Product and Engineering.
• Write tests and evals for every AI feature you build, and monitor them in production.
• Build in human review where people check and approve AI suggestions.
• Follow strict data access and privacy rules in line with Kenya's Data Protection Act, with extra care for data about minors.
• Keep model cost and latency in check.
• Learn our codebase, systems and how schools use Zeraki.
• Share what you build and learn with engineers across Zeraki.
• Ask for feedback and act on it.

• Python
• Java
• LLM API
• prompting
• RAG
• tool use
• SQL
• REST APIs
• Git
• testing
• data cleaning
• data joining
• data analysis
• cloud services
• CI/CD
• machine learning
• MCP
• AI agents

• A degree in Computer Science or a related field, or equivalent practical experience.
• Built at least one project using an LLM API — at work, in study or on your own.
• Familiar with prompting, RAG or tool use, or keen to learn them fast.
• Good grasp of SQL, REST APIs and Git.
• Writes tests and understands why they matter.
• Comfortable cleaning, joining and analysing data.
• Exposure to cloud services, CI/CD, machine learning or MCP (a plus).
• At least one use case you have built using ai agents.