Hey everyone,
I’m excited to share that AI Insights for DHIS2 has been selected as a finalist in this year’s app competition.
After years of watching health workers struggle with complex analytics tools, I built this app to solve a real and recurring problem: making data easier to understand and act on, especially for those on the front lines.
Selecting your Org unit(s) + Data + Period
Chat Interface
Generated HTML Report (Save as a PDF or Print)
Automatically Generated Chart
See It in Action
Watch the short demo video here
https://www.canva.com/design/DAGp6v4aJkw/D5Xzadh43CSxHJbeVPL1hw/watch
It walks through how the app works, what you can ask, and how the responses look with real DHIS2 data.
Why I Built This
I’ve seen how much time goes into building dashboards. We spend hours selecting indicators, polishing layouts, and sometimes even hiring consultants to gather requirements for dashboards and widgets. But once they’re built, the same users still come back asking for the same insights the dashboard was supposed to provide.
Instead of constantly updating dashboards or running repeated training sessions on how to use analytics tools, I wanted to offer a simpler option. With AI Insights, users can just ask their questions in plain language and get immediate answers. No need for specialized training or ongoing support to interpret dashboards. Just a direct way to talk to your data.
What AI Insights Does
With AI Insights, you can ask your DHIS2 data questions in plain English:
- “Which districts are behind on vaccination targets?”
- “Show me the malaria trend for the past 6 months.”
- “Compare nutrition program performance across all health centers.”
You get meaningful answers, charts, and insights without needing any analytics training.
Key Features
Chat Interface
Interact with your data like you’re sending a message. Ask follow-up questions and get context-aware responses. If you can use Whatsapp, you can use this.
Privacy Options
Choose between using OpenAI for advanced analysis or running everything locally with Ollama if data privacy is a concern.
Multi-Facility Analysis
Ask “Which facilities need immediate attention?” and get comparisons across your entire region.
Smart Visualizations
The app automatically generates the right charts and lets you download professional-quality reports.
What Makes It Different
Many data tools are built with advanced users in mind, which can make them difficult to use for health workers who need quick, clear insights. AI Insights is designed to remove that barrier. It supports everyone from clinic staff to program managers by making data more accessible through natural language interaction.
It also addresses some of the common technical hurdles. For example, a built-in proxy helps resolve CORS issues that often come up in hosted DHIS2 environments, making the app easier to deploy in real-world settings.
Try It Out
The app is open source and available on GitHub:
And also via the DHIS2 App Hub: DHIS2 App Hub
To install, upload the ZIP file through App Management in DHIS2.
To try the local AI option, install Ollama and pull a model like llama3. Setup instructions are in the README.
I’d Love Your Feedback
If you try it out, I’d really appreciate your thoughts.
What works well? What could be improved? What kinds of questions would you want to ask your data?
Community Voting
If this app sounds useful for your work, I’d be grateful for your vote here: 2025 DHIS2 App Competition Finalists and Voting
Regardless of the outcome, I plan to keep developing this. The need for simpler, more accessible data tools is too important to ignore.
Catch the Live Presentation
I’ll be presenting AI Insights for DHIS2 during the app competition session at DAC2025 on Wednesday, from 13:20 to 14:20 (Oslo time). If you’re attending, I’d love for you to join the session and hear your thoughts afterward.
Thanks for reading, and happy to answer any questions here in the thread.
John
P.S.
If you’re at the DHIS2 conference, feel free to find me before or after the presentation. I’d love to chat about the data challenges you’re facing and hear how you’re using DHIS2 in your programs.