Ingestion pipeline for RapidPro?

Hi all!

My name is Nic Pottier, I’m one of the RapidPro (https://app.rapidpro.io) developers at Nyaruka and we are exploring how we could get the data we collect into DHIS for our partners.

RapidPro is used by many organizations but most widely by UNICEF in 70+ country offices to do nation scale polling and other interventions. We collect many data points from millions of respondents via SMS, Facebook Chat, WhatsApp and other sources.

We are interested in seeing whether we could build a pipeline to get that data into DHIS which would let our partners use your excellent tools to visualize those results across accounts / countries etc…

That said, we know very little about DHIS2 and what that may look like. Can someone point me to any docs around that (or relevant source code) to point us in the right direction. To be perfectly honest we aren’t even sure whether DHIS2 is the right tool!

This has been a long standing backlog item for us but something which has gained urgency with the Covid19 outbreak, so any help would be highly appreciated. Also happy to hop on a quick call to levelset or coordinate.

Cheers.

Nic and the Nyaruka Team

Hi Nic

We could maybe setup a call if you want to know more about DHIS2?

Regarding our APIs etc you can find documentation:

Please let me know if you have any questions,
Morten
(team integration)

Hi Morten, pleased to meet you!

Thank you for the links to the docs, this definitely put me on the right track:
https://docs.dhis2.org/2.33/en/developer/html/dhis2_developer_manual_full.html#webapi_data_values

I am going to touch base with our partners to better understand which data they want ingested so that I can have an informed discussion with you. After that I will reach out to see if we could hop on a call to answer some questions / clarify assumptions after that.

Thanks again.

Sounds good, let us know if you need any more info. :slight_smile:

Hi @nicpottier
We have explored and done import data from RapidSMS to DHIS-2 ( 3 years ago) in Rwanda,
we can share our experience with that, it was somehow a manual process but it was working fine.
@Muhireandrew and @Randy_Wilson know more about that.