CHAI/VillageReach + Global Fund - Rapid COVID Response

Date: Mar 26, 2020

Participants: @Wesley Brown , Guarav Bhattacharya (CHAI), Iain Barton (CHAI), Mathieu Courtois (Global Fund)

Meetings:

Meeting #1

(Guarav, Iain, Wes)

Notes:

• Limited rollout, specific to COVID response

• Limited products, and configuration

• Quickly spin up instances in the cloud

• OLMIS - Wambo interface

  • Guarav - May be easier to interface with the country warehouse rather than Wambo

  • Ian - Lacking a interface between Wambo

  • For MVP, just focus on interface with warehouse

• Data Ownership

  • Guarav - Ministries should own their data and be able to benefit from it

• What is the max time we have to define and create the MVP

  • Iain is looking for a very fast response on this, having a high-level idea of the MVP by the end of the week

Meeting #2

(Guarav, Wes)

Notes:

• MVP Ideas

  • Limit to COVID Products

  • Stock management and requisitions

  • Include any facilities that might need PPE

    • Should be able to request items

    • Push system (?) where national-level sets the orders

      • National level users can create requisitions for facilities

  • Proof of delivery based on created orders

• Might want to leverage CCE equipment tracking for ventilators

• Ability to get pipeline information (stuff being ordered) in the order fulfillment in OLMIS

  • Or maybe from OLMIS into Wambo for reporting?

  • Perhaps Wambo into OLMIS reporting?

• Setup "Global" OLMIS instance, each country would be a facility in the global instance

  • Automated Import or manual entry of usage data

  • Country requisitions could be created from this global instance and then passed on to Wambo to fulfill

  • The "Buffer Warehouse" is a complication, if this is modeled as a facility then how do country facilities request from it?

• Wambo Integration

  • This could be a lot of work and we're not yet sure how that would look

  • Likely part of a future phase

• Akin is working closely with a dev contractor in Nigeria who would do the work

  • Akin and I would work together to ensure that this work is done well

  • Are marginally familiar with the system

Questions

• How do we simplify the User and Facility setup?

  • Perhaps we could use the WHO list for facilities?

  • Users - ?

Meeting #3

(Guarav, Iain, Wes, Mathieu)

Notes:

Mostly a discussion on how Global Fund can source and procure the PPE and other products.

2 Main options for sourcing:

1. Collaborative sourcing and procurement through wambo

2. Pooled volumes, individual sourcing, and procurement through wambo

Meeting #4

(Guarav, Wes)

Notes: Just further discussion on the MVP and how the buffer warehouse could be implemented in the global instance.

Meeting #5

(Guarav, Satish, Brandon, Wes)

Notes:

Brandon: Might be possible to set up everything within a single instance of OpenLMIS (using the permissions and configuration options)

Need to determine what custom work would be required for the initial MVP:

• Export of orders to Wambo

• Using the CCE service for ventilators

Wambo Integration Meeting (27/3/2020)

(Gaurav, Satish, Mathieu, Kevin (Global Fund), Steven Harsono, @Wesley Brown , @Josh Zamor )

• OpenLMIS integration is normally through flat files

  • Need to have consistent product identifiers - ie, GS1

• What messages would we need to pass?

  • PO (ish), ASN

  • OLMIS has a simplified order that is sends out

  • OLMIS receives a similarly simple file back from the Warehouse/ERP

• Wambo

  • Purchase Requisition is what we would be creating in Wambo (from OpenLMIS)

    • This would then get processed and turned into a Purchase Order by the Wambo system

  • ASN refers to individual PR line items, down to the lot leve

  • Integration does support csv files

  • There are some other country-level fields that would need to be filled out

    • This would likely need to be manually filled out if we're going to this quickly

  • Will need to figure out how to translate between OLMIS and Wambo Product Lists

    • Feed OpenLMIS data directly from Wambo so the data is consistent