AI​/ML Engineer

T-Shaped AI / ML Engineer

Have you ever shipped a complex agentic AI system to production, watched it gracefully (or catastrophically) break under real-world user traffic, and spent the night refactoring its multi-step reasoning layer? If you value production-tested ML rigour over isolated Jupyter notebook prototypes, we have a system for you to build.

About Tiko

Tiko is an African nonprofit committed to strengthening the potential and resilience of adolescent girls across Africa. We address the "Triple Threat" of unintended pregnancy, HIV infection, and sexual and gender-based violence by building local health ecosystems that provide stigma-free, no-cost, quality-assured services.

Our model brings together key local actors: community-based organizations (CBO) with peer mobilisers who act as health companions to girls; public and private health clinics that deliver care; and retail partners that redeem Tiko Miles
-our behaviour-change incentive programme that rewards service uptake and feedback.

We invest in partners by strengthening CBO capacity, training frontline workers and providers, supporting clinic quality improvement, and compensating partners based on performance. Our technology platform connects all actors by enabling referrals, verifying service delivery, facilitating payments, and generating real-time data.

Tiko operates in six countries:
Kenya, Ethiopia, Uganda, Burkina Faso, South Africa, and Nigeria, with additional offices in Portugal, the Netherlands, and the United Kingdom. For a clear overview of our work, we recommend watching this short video.

Globally, our team consists of +250 enthusiastic, international colleagues. Whether you are working from our biggest office in Nairobi, the fast-growing office in South Africa, or from home, our people are young, and our culture is global and dynamic. Our work environment is fast-paced, informal, and friendly.

For this position we will happily accept applicants from South Africa, Kenya, and Portugal.

The Job

We are looking for a T-Shaped AI / ML Engineer to join a high-ownership, agile team tackling one of our most critical scaling challenges: risk and fraud on the Tiko platform. In our engineering ecosystem, "T-shaped" isn't a buzzword- it means you bring a broad, rock-solid foundation in software engineering and data infrastructure (the horizontal bar), combined with deep, hands-on mastery of deploying modern AI and LLM-based systems in production (the vertical bar).

You will move incredibly fast to build, monitor, and optimise intelligent systems that protect our platform. This is a role where you will own entire outcomes from day one, ensuring our architectures remain scalable, secure, and performant for the communities we serve.

Key Responsibilities

Agentic Systems & GenAI Engineering (The Vertical Bar)

• Agent Orchestration:
  Design, deploy, and debug sophisticated agentic AI workflows, focusing on multi-step reasoning, memory integration, Model Context Protocol (MCP), and multi-agent frameworks (using Lang Graph, Lang Chain, or equivalent setups).
• Advanced RAG Pipelines:
  Architect and refine retrieval-augmented generation systems, continually optimising chunking strategies, retrieval design, and end-to-end pipeline evaluation.
• LLM Serving & Optimisation:
  Deploy and operate LLM-based services directly in production, aggressively managing runtime latency, throughput, and cloud costs (using tooling like LiteLLM, vLLM, Sage Maker, or Bedrock).

Production Foundations & ML Rigour (The Horizontal Bar)

• Traditional Grounding:
  Implement robust traditional machine learning models—specifically across tabular/structured data (classification, regression, ranking) or NLP—to solve complex risk and behavioral patterns.
• Production Coding:
  Write idiomatic, clean, and production-quality Python code that seamlessly integrates into our core data and software engineering ecosystems.
• Structured Data:
  Query, manipulate, and pipe structured data at scale using complex SQL configurations.
• Iterative Systems Design:
  Participate in team rituals, maintain meticulous technical documentation, and champion a culture of shared learning and graceful systems degradation.

About You

Yo…