Avni

High quality last-mile service delivery is often a challenge to governments in India. High scale of operations, non-uniform coverage, and diverse needs of each community ensure this by design.

NGOs address these deficiencies by setting up their own last-mile task force. Due to low scale (10k to 100k beneficiaries) and being closer to the community, these NGOs are equipped to identify health hazards faster. They usually already have running programs in their communities.

In our experience, introducing technology to these field-workers has helped reduce workload, streamline processes and improve care delivery. A product for this group must be cheap, easy to maintain, change based on needs, support offline workflows and enable the field-worker.

Avni was conceived to solve these needs has now reached 300,000 beneficiaries and 1M service instances.

By strengthening grassroots organizations, Avni improves last-mile delivery in hard-to-reach places across the world.

Background

Task shifting - outsourcing of well-defined protocols of care to less-skilled individuals is commonly used in LMICs to improve health outcomes with lesser doctors. Typical tasks include maternal care, child-nutrition, adolescent health and hypertension. Indian government employs at least 3 million such workers.

Many NGOs use a network of village health workers (VHWs) to handle primary and preventative care. When pandemics strike, these NGOs can pick up signs early, and manage interventions through their VHW network.

Problem

• At their scale - about 10k to 100k beneficiaries, data management is hard.
• Pen and paper based solutions result in drudgery and a low speed of information flow. Effective management requires timely interventions, especially in a pandemic.
• Many products in this space are focused on specific programs, making them hard to reuse during a pandemic.
• Custom software can be costly to maintain, and can vary in quality.
• Any new software takes time for acceptance.

Need

There is need for a product that

• is easy to set up
• can be modified to suit the needs of an NGO
• is cheap
• works offline
• supports multiple languages
• helps VHWs in their daily tasks
• can be used to quickly respond during a pandemic

About Avni

Avni is an open-source community service and data collection platform.

Avni Project was started in mid-2016 as a open source
community health worker platform designed to aid programs operating
below primary health center level — i.e. sub-center, village and slum
level. It has since matured into a platform that can support most field
work use cases, not just health. Avni is now implemented for water,
social welfare services, educational outcome assessment and health. Avni
(अवनि, અવની, ਅਵਨੀ, அவ்னி, అవనీ, ಅವ್ನಿ) means "earth" in Hindi.

Components of Avni

Features

1. Open Source
2. Cloud-hosted

Impact

Avni has been used to record about 1 million interactions for about 300,000 beneficiaries. Here is a link to our live usage metrics.

Story 1

Shown above is a graph of weight-for-age z-scores (based on the WHO standard) of children below 5 in a cluster of tribal villages. You would notice that the graph is off by about 2 standard deviations. Only 2% of the kids in the village are above a z-score of 0.

This data is from one of the 25 organizations that use Avni. This organization also runs other programs to handle maternity, sickle cell disease, hypertension and mental health.

Before using Avni, getting a graph such as this used to take a few days. It is now easier for them to get an idea of the situation, introduce interventions and look at results much faster using Avni.

Story 2

An organization that deals with adolescent health. This organization performs regular visits to schools and with children in the village if they are not enrolled.

The objective of the program is to look at multiple parameters - demographics, anemia, sickle cell disease, menstrual disorders, early marriage, drug abuse and safe driving habits with the intention of supporting where required, and reduce school dropouts. They work with children between 14 and 16 in the tribal belt of Gujarat. Below is a table of the prevalence of anemia in the area they work in.

Before using Avni, they were using spreadsheets to determine high-risk children. There were missed followups for children in villages and missed high-risk detection in schools. After using Avni, they noticed an increase in the number of high-risk detection. They also found it easier to perform followups.

Other case studies

There are several other case studies available in the Avni website, some of which are being linked below.

Malnutrition project at JNPCT - https://avniproject.org/case-s...

Sickle-cell screening and treatment - https://avniproject.org/case-s...

Maternal and child-health - https://avniproject.org/case-s...

How we engage with organizations

Many implementations require us to travel to the area of operation to understand their needs before suggesting and implementing a solution. We also try to perform a training (between us, we can cover 7 languages, but if it is not possible, we train a trainer who speaks a common language such as Hindi or English and let them train the VHWs).

We take their paper records, or spreadsheets to get an idea of the operations, then implement them with Avni, sometime with modifications to improve their processes.

VHWs are usually semi-literate, they need significant training for any technology. To respond to an emergency, we need to take out friction of technology. 

You need a tool that you use for your daily work that can easily be adapted to any new need in the community for successful adoption. We see this happening today.

We have noticed many existing Avni users roll out Covid programs on Avni on their own. One of them interestingly implemented a call-center for daily check-ins for low-risk patients. Another started a survey to identify needs of Covid affected families.

In the past 5 years, we have had 25 organizations adopt Avni.It has served about 387,000 beneficiaries.

We believe they are happy using the system because

1. These organizations have been renewing their contract every year.
2. We started the product using grants, but currently, all development of new features are being funded by existing customers.

Here are the issues with software solutions in this space

• Some solutions focus on specific programs, while a community worker has to deal with many of them (eg: Anmol ).
• Some focus on data collection, not so much on acting as a work-aid for the community worker (eg: ODK).
• Some software is good, but are based outside of India or focused on bigger programs, therefore costly (eg: Commcare).
• The price-point for the customers we work with can only be achieved with significant funding on the original software development, which explains the lack of serious software adoption in this space for this customer segment.

We intend to solve this through

• Using grants for majority of the core software development to reduce cost to customer
• Empower end-users to take control of their programs through simpler interfaces. We have seen this happen with many existing implementations.

• Encourage other software organizations to consider using the hosted Avni
  solution. This brings in orgs without hosting experience to provide
  low-code and cost-effective solutions to NGOs. We already have 2 implementations managed by a different software provider.
  

• Keep the solution open-source to encourage other software providers start Avni hosting services on their own. Nobody has done this yet, but we have not tried looking either.

Currently serving - 387,000

In one year - 750,000

In five years - 6,000,000

Outputs of Avni have been mentioned in the previous question's answer.

Measuring outcomes depend on our customers' work where we only play an enabling role. All credit to achieving outcomes, especially when mapped to UN sustainable goals would go to the customers.

Samanvay is the organization that works on Avni. Samanvay employs 9 full-time staff.

We have a fluid structure where anybody can be assigned to Avni based on need. At any given point in time, about 4-5 people will be working on Avni.

Most of the team working on Avni have had significant prior experience working on open-source products or for the development sector. Some examples of products that we have built or contributed to include

• Bahmni - An open-source hospital management used in 400+ sites and 50+ countries
• Maatritva - A maternity high-risk detection solution used in parts of Maharashtra, India

During our time in Samanvay, we have also developed other products in the development space. Details are available in the website.

During the course of our previous experience and during Samanvay, we have regularly been doing field visits understanding field-work. One of our employees has experience of implementing and running quality improvement programs in primary-care clinics in tribal districts of Madhya Pradesh, India.

Most of us have traveled extensively to these places as part of these field visits. Avni was conceived at one such place to enable healthworkers prescribe simple medications for fever, malaria etc in a tribal village that gets cut off from the clinic during monsoon floods.

At Samanvay, everyone is considered an owner of the company. Anyone can suggest projects to start, choose their area of work (based on capabilities) and contribute in their own ways.

Employees come from 6 different states of India.

Out of the existing employees, 2 out of 9 are women. It is hard to change this because of our speed of growth. Our last hire was more than a year ago. Also, most of our hires were inbound calls of passionate people.

We were trying to see how we can speed up our product development so that we can provide more value to those using Avni. 

We informally get advise from our earlier partners, and have found it useful. We believe having more people to bounce ideas off will help us grow the product and improve impact in places we have not thought about before.

Everyone in our team has a technology background, with some experience in the development sector.

Sales in this sector is a long process (even small deals take a long time to convert). All our sales have come in through our networks and referral only. We have not found good ways to reach out to prospective customers. One of the things we tried was running monthly webinars, we've had some minor success there, but not too much. Some direction in creating a marketing strategy can help.