Making AI work for NTD programme leaders

AMPLIFY 2027 is an eight-session course for staff of national NTD programmes. You bring one real task from your own work. Every session sharpens it or builds the AI skill to solve it, and you finish by presenting the solution.

Learn AI on a problem you own

Leave with a working AI solution to a real task from your NTD programme, and the skills to keep it running after the course ends. Eight live sessions over four months, free for national programme staff.

Click any session to see what it covers and what you do before the next one. Interactive sessions also open a hands-on learning module.

Your problemat the centre of every session

Setup, AI fundamentals & prompting

Session 1 · Foundations

You meet the cohort and the instructors, and leave with your own Claude project and a place on the course Slack. We introduce the vocabulary you will use for the next four months: what the different models are good at, the red lines on what must never go into an AI tool, and the habit of judging every answer, then keeping it, trying a new idea or discarding it. From there we get straight into prompting: how an effective prompt is put together, and how to get answers that sound like you rather than a chatbot.

Before the next session, you introduce yourself on Slack, post your problem for the whole cohort to read, and respond to at least two other people’s. By the end of the week your problem has a name, and other people are already thinking about it.

Getting started with Claude TeamsJoin the AMPLIFY workspace, take a tour of the app, and learn to share projects and artifacts, and what stays private to you.

Start here

Interactive learning moduleChecklists, a prompt builder and exercises for this session, in English and French.

Open session 1 module

AI training is widely available online, but most courses are general-purpose and use examples that have little to do with NTD programme work. Vendor-built tools have helped national programmes tackle specific challenges with AI, though these are not always the national priority, and tools delivered through short-term projects have often been difficult to maintain once that support ends. AMPLIFY offers training and mentoring that is specific to NTD programmes and their data. Courses are taught with nationally reported data available from WHO, and focus on the skills NTD leaders need at country level to build and maintain solutions long after the course ends.

Start with the task, not the AI

Your application begins with a problem statement. It asks about your work, not about AI: one specific task you do again and again, described as if a new colleague were taking it over next month. Everything you build in the course grows out of it.

  • Data management and analysisYou compile, clean, check or analyse programme data, whether it is numbers or interview notes.
  • CommunicationYou write reports, prepare slides, or run social and behaviour change campaigns.
  • Knowledge managementYou bring information together from different places and get it to the people who need it, when they need it, including through training.
  • Project managementYou plan, track and coordinate the work itself.

A weak problem statement

“Data management is a big challenge in our programme. Reports are often late and there are many errors. AI can help us improve data quality and efficiency.”

Which data? Which report? What errors? How late? Nobody could design a project from it.

A strong one

Task
I compile district MDA treatment data into a presentation, with charts and maps, for the national LF and oncho technical meetings.
How often
Once a year. It takes me and a data officer about three weeks.
Last time
After the 2025 round, 38 districts sent their summaries: 25 as Excel files by email, 9 as WhatsApp photos of paper forms, and 4 by phone. Several used old versions of the template, and I retyped every figure into our national workbook.
What went wrong
One district reported more people treated than lived there, which gave 112% coverage. I only noticed when I made the chart, after the numbers had already gone to a partner.
Better looks like
Every district’s data checked for obvious errors within a few days of arriving, and the report finished in one week instead of three.

Illustrative example, abridged from the AMPLIFY problem statement form.

AMPLIFY Cohort 1 (2026): sixteen NTD professionals

Five students from Cohort 1 will return as peer mentors for Cohort 2!
Cohort 1 → Cohort 2
We’re excited to see participants supporting each other.

Our first cohort ran from March to September 2026: seven monthly modules in English and French, built entirely on the free tiers of AI tools. Each participant was nominated by their national NTD programme to be its AI focal point.

  • Burkina Faso1
  • Côte d’Ivoire2
  • DR Congo3
  • Ethiopia3
  • Ghana3
  • South Sudan2
  • Togo2
  • Seven countries16
Slide recapping the P-T-C-F-I prompt frame: persona, task, context, format, iterate, with an NTD example prompt
The P-T-C-F-I prompt frame
Slide showing the refinement loop: ask, evaluate, identify the weakest element, refine
The refinement loop
Slide on descriptive statistics with a bar chart of treatment coverage by district
Descriptive statistics on coverage data
Slide for the Screenshot and Explain exercise, using a chart from the ESPEN portal
Screenshot & explain, on ESPEN charts

Slides from Module 2, Prompting & Data Literacy, April 2026.

AMPLIFY Cohort 2:
Session 1, winter/spring 2027

Enrolment for Cohort 2 will be a competitive process for invited country pairs
Applications now open
Cohort 2 is made possible through the generous support of the Gates Foundation and Sightsavers.
Sessions
8live, online
Cadence
Every 2 weeksabout 4 months
Languages
EN · FRthis round
You apply as
A pairsame country
Cost for AMPLIFY
Freeall fees are covered

Who should apply

  1. RoleYou work for a national NTD programme in a Ministry of Health.
  2. ExpertiseDeep NTD knowledge: you know the programmes, the data and the decisions.
  3. DriveYou are motivated to use agentic AI on real programme challenges.
  4. PairPriority goes to two applicants from the same country who will learn together.

How you'll learn

  • Every session is build time. Short teaching, then you work on your own real task with live help.
  • Homework between sessions, submitted and discussed with the group on Slack.
  • Feedback on what you made, in pairs and small groups.
  • Mentors from Cohort 1. Five alumni return to coach the new cohort.

This round runs in English and French. More languages will follow in a second training in Q3 2027.

Apply for Cohort 2

Space is limited · committed participants only

Attendance and assignments are mandatory. This is a build-along course, not a webinar series, so apply only if you have the time to commit.

Student profiles — coming soon

AMPLIFY: Short courses

This section provides access to a growing list of short courses that are intended to cover advanced topics or highlight partner tools. If you have a short course or a tool you want highlighted, please reach out to apavluck@sightsavers.org and we will cover it here.

  1. Preview of the chart-chooser short course

    20 min Beginner

    Choose the right visual

    Not sure which chart to use? Answer two or three quick questions about your data and the interactive chooser recommends the right one, shows a real example, and even writes the prompt to build it — plus a cheat sheet you can reuse every time you brief your programme.

  2. Preview of the Claude Teams short course

    40 min Beginner

    Claude Teams: work together, keep what's yours

    Join the AMPLIFY team workspace and learn to work as a group: build a shared project with common files and instructions, publish an artifact your whole team can open with one link, and play a quick game to learn exactly what teammates can see — and what stays private to you.

  3. Preview of the Subscribe, rent or run your own AI short course

    30 min Intermediate

    Subscribe, rent, or run your own AI?

    A frontier subscription, open-weight models paid per token, or a free model on your own computer: what each costs, what it can do, and who sees your data. Answer three questions for a recommendation, then compare the monthly cost for your own workload.

  4. Preview of the GitHub + Vercel short course

    45 min Intermediate

    GitHub + Vercel: put your work online

    Get your work on the internet, free. Set up GitHub and Vercel, publish a live site at your own .vercel.app address, then hand the pushing to your AI agent — you ask in plain language and it commits and publishes. Windows and Mac covered, with a troubleshooting FAQ when things go sideways.

  5. Preview of the OpenRouter + OMP short course

    60 min Intermediate

    OpenRouter + OMP: from chat windows to terminal agents

    One free account, every major AI model. Try Claude, GPT, Gemini, Llama and more from a single place — free-tier models cost $0, and when you're ready to scale you can see exactly what each session costs. You'll finish by turning real programme data into polished slides, straight from your terminal.