How to apply for (Associate) Director, Program Innovation & Evidence, The AI Access Initiative

About Evidence Action

Evidence Action scales cost-effective, evidence-based programs that reduce poverty across Africa and Asia. The AI Access Initiative (2AI) is incubated at Evidence Action and focuses on scaling AI-enabled 'big bets' for people in low- and middle-income countries. It is led by former Evidence Action CEO Kanika Bahl and advised by Nobel Laureate Michael Kremer and Anthropic CEO Dario Amodei.

About the role

The (Associate) Director, Program Innovation & Evidence will build and lead 2AI's evidence strategy for AI in Agriculture and AI in Health programs. This role sits at the intersection of global development actors, top AI labs, and leading researchers, and is responsible for creating open-sourced playbooks, toolkits, and insights that define how to design and launch tractable and impactful AI-enabled programs. The work directly shapes how AI benefits reach tens or hundreds of millions of people in poverty.

A typical day

You might spend the morning reviewing evaluation results from an AI agriculture pilot and the afternoon on a call with an AI lab and a research partner to design the next phase. You could also draft a section of a public playbook or work with the team to define metrics for scaling. The exact daily routine is not specified in the job details, so ask about team structure and typical workflows in your interview.

Who Evidence Action is looking for

• Experience designing and running impact evaluations or evidence-generation programs in low- and middle-income countries, ideally in agriculture or health.
• Familiarity with AI/ML applications in development contexts, including working with AI labs, researchers, or technology partners.
• Track record of translating research findings into practical tools, playbooks, or implementation guides used by others.
• Comfort operating in a startup-like, fast-moving environment with evolving priorities and a small team.

Tips for this application

• In your resume, highlight specific programs where you generated evidence that changed implementation or policy in LMICs.
• Name the AI labs, research institutions, or development actors you have worked with directly, since 2AI operates at that intersection.
• Show how you have scaled a program or idea beyond a pilot, because 2AI focuses on scaling 'big bets' to tens or hundreds of millions.
• Reference the two starting programs (AI in Agriculture and AI in Health) and explain how your background fits one or both.
• Mention any experience creating public goods, open-source tools, or playbooks, as that is a core output of this role.

What to cover in your cover letter

['Your direct experience generating evidence for programs in LMICs and how you used it to improve or scale those programs.', 'Specific examples of working with AI labs, researchers, or technology partners to deploy AI in development settings.', 'How you would approach building evidence for AI in Agriculture and AI in Health, given the fast pace of AI and the slow pace of development systems.', 'Why you want to work at 2AI specifically, including its incubation at Evidence Action and its leadership and advisors.']

Research before applying

• Read Evidence Action's published work on cost-effective programs and their evidence standards.
• Look up Kanika Bahl's background and her role on Anthropic's Long-Term Benefit Trust.
• Review Michael Kremer's work on development economics and randomized controlled trials.
• Search for existing AI in agriculture and AI in health projects in LMICs to understand the current landscape.

Likely interview topics

Based on the job description, expect questions about:

• How would you design an evidence strategy for an AI-enabled agriculture program in a low-income country?
• Describe a time you worked with an AI lab or research partner to deploy a technology in a development context. What worked and what did not?
• How do you balance the need for rigorous evidence with the need to move quickly when AI capabilities change every few months?
• What does a 'public good' playbook look like for AI in health, and how would you ensure it is used by others?
• How would you measure whether an AI program is reaching people in poverty at scale?

Common mistakes to avoid

• Focusing only on AI technical skills without showing how you have generated evidence or scaled programs in LMICs.
• Treating this as a traditional development role and ignoring the fast pace and uncertainty of AI capabilities.
• Writing a generic cover letter that does not mention 2AI, its two starting programs, or its leadership and advisors.