AI Solutions Architect

Code for Africa (CfA) is a non-profit organization that operates the TechLab, which includes the AI Sandbox. CfA is part of the Plus Alliance, a non-profit partnership working across Africa, the Middle East, Europe and the US.

This role involves designing, shipping, and maintaining AI-enabled tools and workflows for teams across the Plus Alliance, setting patterns for others to build upon. It is a hands-on position requiring coding, integration configuration, user testing, and support for deployed solutions. The initial focus will be on Business Development workflows, specifically proposal and partner-intelligence, with future expansion into research, editorial, investigative, and programme operations. The architect will leverage existing tools such as Google Workspace, Slack, Airtable, and CRMs, alongside in-house capacities like DataLab for data engineering and scraping, iLAB for AI-driven forensics, CivicSignal for NLP, and the broader AI Sandbox team for shared model access and infrastructure. This opportunity allows the successful candidate to shape the alliance's use of AI, work with real users from day one (including Business Development, forensic researchers, journalists, and programme staff), and take ownership of design decisions within established architectural and security boundaries, supported by dedicated in-house data and AI teams.

1. Ship tools, workflows, and lightweight interfaces end-to-end, covering discovery, design, build, evaluation, rollout, training, monitoring, and handoff.
2. Prioritize the simplest effective solutions, utilizing models or agents where language interpretation or ambiguity handling adds value, and plain deterministic code or rules for frequent, consistent, or auditable tasks.
3. Treat AI output as unverified until checked, implementing evaluations, source traceability, automated checks, and human oversight for critical decisions.
4. Build the function beyond just tools, establishing an intake and prioritization process, a reusable library of templates, connectors, and review components, and clear rules for central versus programme ownership.
5. Ensure maintainability through version control, code review, comprehensive documentation, and adherence to shared Plus and AI Sandbox platforms and security standards.
6. Maintain cost visibility by estimating running costs, comparing managed, open-source, and in-house options based on cost, security, portability, and maintenance, and replacing expensive SaaS where a clear case exists.

1. A bachelor's degree in computer science, software engineering, information systems, data science, or a related field, or equivalent practical experience.
2. At least 3 years of experience building and deploying internal tools, automation, or AI-enabled workflows that are regularly used and maintained post-launch, demonstrating comfort in navigating ambiguity to deliver working systems.
3. Hands-on experience building with hosted model APIs and open-weight models, including structured outputs, tool use, agents, and integration protocols like MCP, along with familiarity with AI coding and workflow tools such as Claude Cowork, Claude Code, or Codex, and understanding when not to use them.
4. Integration experience with several business systems, such as Google Workspace, Slack, Airtable, CRMs, or email, including APIs, webhooks, OAuth, scoped credentials, and rate limits, with a default to live connections and knowledge of how to prevent cached or synced copies from being mistaken for current data.
5. Practical AI and data risk awareness, including identifying confident but incorrect outputs, handling prompt injection, protecting Personally Identifiable Information (PII) and confidential data, and matching human review to the potential cost of an error.
6. The ability to map actual workflows, design solutions collaboratively with non-technical colleagues, and ensure adoption through effective guides.
7. Strong engineering discipline, including Git, code review, relevant testing, and clear documentation.
8. Preferred: A postgraduate qualification in artificial intelligence, machine learning, or a related field.
9. Preferred: Experience in non-profit, media, civic tech, or international development, including proposal or partner-management workflows.
10. Preferred: Experience with document-heavy pipelines, such as parsing PDFs and DOCX files, structured extraction, and retrieval over frequently changing knowledge bases.
11. Preferred: Formal evaluation practice, including test sets, regression evaluations, red-teaming, and output validation.
12. Preferred: Experience with managed model services and gateways, such as Amazon Bedrock or OpenRouter.
13. Preferred: Cost modeling for Large Language Model (LLM) workloads and security basics like secrets management and vendor security review.