AI Solutions Architect

Role Overview

As the alliance’s first dedicated builder of AI-enabled internal tools, you’ll turn teams’ day-to-day needs into dependable workflows, integrations and lightweight software. You’ll begin with Business Development proposals and partner intelligence, then extend your work to research, editorial, investigations and programme operations, helping colleagues spend less time on repetitive tasks and make better-informed decisions.

You’ll take responsibility from discovery through launch and ongoing support, working within established security and architecture boundaries. You’ll build on the alliance’s existing platforms and collaborate with its data engineering, digital forensics, NLP and AI infrastructure teams rather than duplicating their work.

Key Responsibilities

• Work with Business Development colleagues to identify high-value opportunities and build production-ready tools for proposal preparation and partner intelligence; expand to other teams as priorities are agreed.
• Deliver complete solutions, from understanding a process and designing an approach to development, testing, user onboarding, monitoring and handover.
• Choose methods to match the task: use AI when language interpretation or ambiguity requires it, and favour predictable, auditable code and rules when consistency is essential.
• Make AI-assisted workflows safer and more dependable through evaluation, source traceability, automated validation and human review for consequential decisions; account for inaccurate or malicious external content.
• Establish a shared way to receive, assess and prioritise requests, and create reusable templates, integrations and review components for teams across the alliance.
• Maintain clear, usable documentation and apply version control, code review and appropriate testing so colleagues can support and extend what you build.
• Work with the CTO, programme leads and internal technical teams to ensure solutions fit shared platforms, security requirements and existing systems.
• Track operating costs and assess managed, open-source and in-house options for affordability, security, portability and ongoing maintenance, including potential alternatives to costly SaaS products.

Requirements & Qualifications

• A degree in computer science, software engineering, information systems, data science or a related discipline, or equivalent practical experience.
• At least three years of experience building, launching and maintaining internal tools, automation or AI-enabled workflows used regularly in real operations.
• Practical experience with hosted model APIs and open-weight models, including structured outputs, tool use, agent-based approaches and MCP or comparable integration protocols.
• Familiarity with AI-assisted coding and workflow tools such as Claude Cowork, Claude Code or Codex, alongside sound judgement about when conventional development is the better choice.
• Experience connecting multiple business platforms, such as Google Workspace, Slack, Airtable, CRM systems or email, using APIs and webhooks, with an understanding of OAuth, scoped access and rate limits.
• Ability to keep integrations grounded in current source data and prevent cached or synchronised records from being mistaken for live information.
• Working knowledge of AI and data risks, including unreliable outputs, prompt injection, personal and confidential information, and setting human oversight according to the consequences of errors.
• Strong process-discovery and collaboration skills: able to understand how non-technical colleagues work, design practical solutions with them and help them adopt new tools.
• Engineering fundamentals including Git, code review, appropriate testing and clear technical documentation.

Preferred: Postgraduate study in AI, machine learning or a related area; experience in non-profit, media, civic technology or international development; or familiarity with proposal and partner-management processes.

Additional useful experience includes extracting structured information from PDFs and DOCX files, building retrieval systems over frequently updated knowledge bases, running formal evaluations such as regression testing and red-teaming, and using managed model services or gateways such as Amazon Bedrock or OpenRouter. Cost modelling for language-model workloads, secrets management, vendor security reviews, and creating shared standards or components adopted by other teams are also advantageous.

What We Offer / Why Join

• A founding role with the opportunity to shape how AI tools are built and used across the alliance, including standards other teams can reuse.
• Access to real users and operational needs from the outset, beginning with Business Development and potentially extending to researchers, journalists and programme teams across Africa and beyond.
• Support from in-house data engineering, AI forensics, NLP and shared AI infrastructure colleagues, with ownership of implementation decisions within agreed technical and security boundaries.
• A first-year focus on delivering four to five operational tools or workflows with monitoring and named owners; establishing an alliance-wide request process; and creating reusable patterns adopted by at least two teams.
• The opportunity to use evidence from incoming demand to recommend how the AI hub should develop, including the potential shape of a team you may later lead.