Crowdsourcing testimonies from the field

Sometimes adversity is the perfect catalyst to invent what was unforeseeable in normal circumstances. The post-election violence in Kenya at the beginning of 2008 prompted a bunch of programmers to create an open-source platform that would collect eyewitness reports of violence sent in by email and text-message and map that information onto a Google Map for everyone to see. What would have otherwise gone unnoticed to the world at large became a visual testimony to the violence.

The platform, Ushahidi – which means “testimony” in Swahili – since then has been deployed for a number of crisis and disaster situations to provide critical and life-saving data during emergencies especially when mobilizing relief and rescue operations as in the case of Haiti or the most recent Christchurch earthquake in New Zealand.

Video: What is the Ushahidi Platform?

As with other areas, being open-source, the potential for its application in the context of rural development and financial inclusion are manifold. For example an organization serving a remote rural population can deploy it to collect and map customer feedback about its product or services on a geographic basis, and effect corrective actions based on the different clusters on the map. Government schemes could be mapped and individuals in villages could be encouraged to report any leakages; a surge of dots over a particular village can guide authorities in pinpointing localities that require attention. In the context of MFIs, those dots could come to visually represent repayments by borrowers or perhaps in the current environment as a feedback against alleged harassment by recovery agents.

While the platform per se is crucial and the ways in which it could be deployed are left to one’s imagination, technology however is just 10 percent of the solution, with the other 90 percent being educating people about the deployment and getting them to submit content.

In a short Q&A with Patrick Meier, Director of Crisis Mapping & New Media, Ushahidi, we discuss the platform and its applications.

From its beginning in early 2008, how has Ushahidi evolved, and with it the whole crowdsourcing phenomenon?

The Ushahidi platform has become easier to use (see Crowdmap.com) [Crowdmap is the easy-to-install, hosted version of Ushahidi – think of it as the equivalent of WordPress.com blogs for WordPress] with more features and functionalities. We have also moved to a plugin architecture, which means that third parties can develop their own apps. This makes the platform a lot more extensible and versatile. The application of crowdscourcing to crisis environments was in large part spearheaded by the Ushahidi platform.

The wisdom of the crowd is what an Ushahidi deployment feeds on. What are the inherent challenges in this and how could they be addressed?

I’m not sure it’s as much the wisdom as the presence of the crowd. The crowd will always outnumber the expert journalist, the expert election monitor, the expert disaster responder. There are tradeoffs, more information, but one of the main challenges is validity of that information. But the more information you do get, the greater the probability that you can triangulate said information, ie, the great the chance that more than one person will report about the same event. That’s why we’ve developed Swift River, to help triangulate and verify crowdsourced information. The Ushahidi platform also allows for multi-media reporting, which allows the crowd to share pictures and video footage of an unfolding situation.

Crisis mapping is what predominantly Ushahidi is associated with – Whether it be for Haiti, Snowmaggaden, Australian Floods, Sudan or the recent Egyptian protests. I am sure there are more areas in which the platform could be deployed than just for crisis situations?

That’s correct, the majority of Ushahidi platform applications are not crisis related. The platform is simply a free and open source tool that facilitates collaborate live mapping. The Ushahidi platform has been used for citizen journalism, local governance, environmental monitoring, ICT4D, etc.

Are there instances where it has been used successfully to track progress of rural projects or to track down the lack of basic services on a geographic basis?

Yes, the Kiirti project in Bangalore.

In the Indian context, where a majority of the population lives in rural India, how can Ushahidi be used for rural development?

We’ve developed an audio mapping component that allows illiterate populations to share what they’re reporting via voice. They can leave short voicemail reports that can then be mapped or transcribed.

How can companies or NGOs operating on the field, especially in remote rural areas, benefit from it and information filtering tools like Swiftriver?

Filtering tools like SwiftRiver are more useful for events that have a lot of main/social media coverage. For remote rural areas, the audio mapping component may make the most sense. Also, one can use FrontlineSMS for offline SMS mapping, ie, you can do on site mapping and then upload the data to the Ushahidi platform.

You have your Crowdmap Check-In service slated for launch at SXSW 2011 which gives individuals and organizations the ability to create their own “Foursquares” or “Gowallas” if you like. What are your expectations from it?

We expect that CI will further lower the barrier to mapping and we thus expect more live maps to be created as well.

Where do you see the project in 2-3 years time? And in general the areas that need to be worked upon when it comes to crowdsourcing information from the field?

It’s hard to say where the project will be in just 1 year because things are moving so fast. I think the area that needs most work on is SwiftRiver, ie, improving our tools to curate and validate crowdsourced information in real time.

About Partick Meier: He is the co-founder of the International Network of Crisis Mappers and previously co-directed Harvard University’s (HHI) Program on Crisis Mapping and Early Warning. He has an MA in International Affairs from Columbia University and is currently pursuing his PhD at the Flectcher School of Law and Diplomacy.


Technology to the rescue of MFIs

– By Sameer Segal, Founder & CEO, Artoo

[Artoo Slate is a software solution designed for microfinance field staff that takes the entire process of data collection and loan disbursement online. Sameer has been recognized as one of Asia-Pacific’s most promising young social entrepreneurs by the Paragon100 Fellowship. He holds a B Tech from the National Institute of Technology, Karnataka and is a StartingBloc Fellow (MIT Sloan). His passion is inclusive technology, something he discovered during his internship with Ujjivan. He, along with Co-founder of Artoo, Indus Chadha, has developed Artoo Slate which helps microfinance companies cut down on operation costs. He shares with us in this guest blog, how technology can make a difference to microbanking institutions that cater to the bottom of the pyramid.]

The Malegam Report is finally here. At first glance, we were all glad to see how well balanced it appeared. But now we have to begin to make sense of the constraints that it places on MFIs in the short term. In the words of Vijay Mahajan of BASIX: “some provisions are so severe that some MFIs will be facing death by April”.

Indeed, most MFIs must be grappling at the moment with what changes they will need to make to stay alive.

If MFIs need to reduce costs, remove redundancies, and improve efficiencies at all levels, they need to centralize their operations. Centralizing will also help MFIs ensure the quality of the customers they acquire and thereby reduce risks. To centralize operations and still maintain competitive TATs for all customer-centric activities (e.g. customer acquisition, loan disbursement, and repayments) is the hardest part of the puzzle that needs to be cracked. It will be only possible for MFIs to consolidate branches and have their field agents operate over larger geographies (improving borrower to employee ratio) when they can monitor and remotely manage their staff and activities. For that they need technology. 

We believe, however, that this needn’t be a question of the survival of the fittest. It could serve as an opportunity for visionary MFIs, regardless of their size and strength, to re-imagine their operations in a way that, while respecting the RBI’s imminent mandates, dramatically reduces their Operating Expense Ratio (OER) and enables them to remain profitable and survive in the face of the Malegam Report.

At Artoo, we wish to catalyze development through inclusive technology and empowering communication. Our software framework, Artoo Slate, can help MFIs bring down their OERs to meet the RBI’s requirements in a timely manner while enabling them to remain profitable. We believe it has the potential to help MFIs become more productive in helping their customers rise out of poverty. Here’s our take on what the Malegam Report is asking MFIs to do and how we might be able to help.


 Artoo Slate is a software solution that takes the entire process of data collection (under 18 minutes for complete customer acquisition process*) and loan disbursement online (70+% of Loan Applications can be processed in the field on the same day*). It will capture rich data from the field, do away with the back and forth of paper, avoid innumerable delays (reduction in turn-around-time (TAT) from 3+ days to 1 hour*), and drastically reduce expenses (courier, Document Management System Hubs & outsourced data entry). It will allow for easy exchange of data between field staff and backend systems (CBS/MIS) in a way that will reduce time spent (41% of center meetings take under 1 min to update paper work) on customer query clarification and identification and resolution of errors in customer profile and loan application forms. Even while the credit bureau is stabilizing, it will enable MFIs to implement a field credit check upfront for renewal loans based on internal data and assessment. 


Our framework enables field agents to operate remotely and helps distributed MFIs to centralize their operations, while improving their TAT for all customer-centric activities. MFIs can monitor their business on a real time basis: pick up on trends (mass default, political/economic turbulence) as and when they happen directly from the field (defaulter information available instantaneously as compared to 10-15 days lag in previous implementation*). In addition, MFIs can track their social performance on a daily basis.

It is an intuitive interface that has been designed keeping in mind field staff’s educational training and exposure to technology. It will also serve as platform through which MFIs can train (e.g. basic English skills, computer skills, updates on new products and offerings) their field staff on-the-go and monitor them on a real time basis to improve their overall service quality. MFIs can improve their field agent quality and build their capacity, reducing their attrition to short-term focused aggressive competitors.


Artoo Slate, in the hands of the field agent, promises to be a scalable way for the MFI to engage more effectively with their end customers through videos, graphics, and other interactive media (imparting life skills, financial planning, healthcare information, conversational English, etc.) Engaging with the end customer will not only give them a reason to attend center meetings but also allow them to recognize their MFI as a real partner in their struggle to climb out of poverty, giving forward-thinking MFIs an opportunity to differentiate themselves, improving customer loyalty and therefore profitability.


We have been really lucky to pilot our solution, Artoo Slate, at Ujjivan microfinance, and are happy to share the interim results of our pilot here. The pilot covers a branch in urban Bangalore and includes processes of customer acquisition, collections, branch transactions, and field agent training.


Leveraging Training through Technology

By Chandrachudan V and Rajesh E, IFMR Rural Finance

For a business operation, a great deal of what happens on the field is determined by how equipped its field managers are. In our case, our Wealth Managers (WMs) at KGFSs (Pudhuaaru, Sahastradhara, Dhanei) are the primary interface in our endeavour to ensure access to finance.  Complete, continuous and accessible training therefore is a crucial ingredient in sharpening their skills at all times.

While the current training process involves a 24-day induction training and regular schedules of refresher training, this trainer-led effort with a lot of manual intervention provides a lot of challenges especially as it involves a lot of paperwork. To top this, there is no centralised repository that the WM could be directed to in order to stay updated on happenings and changes relating to products offered, processes and our USP of Wealth Management.

To bridge this gap, we hit upon the idea of an online learning management system which would be a one-stop shop for all learning needs featuring – product modules, process modules, Wealth Management modules, e-test and e-certification, storage of all training related information of each employee which later on can be factored into Performance Management of the respective employee.

Importantly this centralized learning environment will ensure consistency among learners, thus promoting web–based training that encourages self and participative learning, thereby, reducing trainer’s intervention for all training programs.

From ideation to fruition, it involved a lot of research in finding a suitable platform that could be customised to suit our training needs. The search was on to finalise a platform that is robust, user-friendly, easily customisable, secured, is being widely used across top companies with proven records and also a platform which is compliant to training standards like SCORM.  After 6 months of research, we decided on “Moodle”- a popular and prominent, free and open-source e-learning software. Moodle’s ease of installation, features and the level of customisation that could be done to it made it a perfect fit for our needs.


Learning Management System Architecture

Customised into 3 local languages (Tamil – Pudhuaaru KGFS; Oriya – Dhanei KGFS; Hindi – Sahastradhara KGFS) and in English, the learning management system would be a comprehensive tool that in its first phase would provide:

•    E – Learning on all products/processes and concepts of Wealth Management
•    E – Certification for all products/processes
•    Scheduling of training programs and maintaining repository of training data
•    Quiz and score management
•    Training dashboard automation: online extraction of training data – employee/geography wise
•    Automated mailer notifications for users
•    Feedback automation with effective reporting
•    Video based training on key aspects for better understanding
•    Internal chat for clarifying queries instantly
•    Discussion forums for knowledge sharing

Pic2_LMS Pic3_LMS Pic4_LMS
Customised homepages of KGFSs (Click on image to enlarge)

The system would make it mandatory for the WMs to pass through the different certification programs on the various products, processes and concepts of wealth management. Also the system would update the training manager/CEO of a particular geography with macro and micro level information as regards the training status and the needs of the local staff so that they can evolve their training efforts.

The system will be live soon with the above-mentioned features. Improvisations and enhancements are planned in Phase II that will also include animated learning on all products/processes and concepts of Wealth Management, which will supplement the current modules.


There’s a Map for That

The power of data is in being able to make sense of it. Sometimes however, under the enormity of data, the best of systems can throw a challenge in aiding the right decision. Hence there is a need for a comprehensive visual component to complement existing systems, which would allow the absorption of large amounts of data in a presentable format. Thus is an effort by IFMR Rural Finance team in collaboration with CDF to add a geospatial dimension to the existing MIS.

The new system aims to feed in the enormous amounts of data of KGFS customers and presents their details in a visual format laid across over an interactive Google map. With seamless integration with the existing databases of KGFS, the system would provide real time and most immediate information on the dot.

Some salient features of the Geographic Information System are:

  • Ability to perform basic descriptive analysis (count, mean, variance, etc) on the map at the macro level
  • Answering queries across various parameters. For eg. Transaction frequency of customers in a particular village, Loan Outstanding (<5K, 5-10K, 10-15, etc.) village-wise, insurance customers with 2 KM radius from branch location, non-customers in hilly regions over X feet above mean sea level, etc.
  • Ability to drill down to the customer / transaction level

Being developed with the help of open source tools like Google Maps API, PostGIS and MapServer, the power of the system would be made available to every constituent of KGFS from the Wealth Manager to the CEO. Access to some features would also be made available on the public website of IFMR Rural Finance.

Early snapshot of the system showing the Karambayam Branch Service Area:


Zoomed-in map showing a smaller area


Ajay Karthik and Amit Shah who are involved in the development of the system say, “The visual representation of information on the map is expected to be more intuitive and hence at least one notch above the most powerful MIS. The system will help in providing insights into geographical gaps in the market / service area. It is also expected to help in taking decisions on branch locations, optimisation of distance from HQ / service area from a logistics point of view.”

Watch this space for more updates while the system is being developed.


Technology Migration at KGFS

Recently our Rural Finance Technology team migrated our legacy core banking system to a new core banking system.

Typically the migration involves three stages: Pre, Intermediate and Actual migration. At the pre-migration stage, a great level of attention went into understanding the data and the various parameters that hold it. At the intermediate stage, a mock-migration under a controlled testing environment was performed to iron out data inconsistencies and error incidents.  The actual migration, that happened over the weekend so that systems are in place when the week starts, happened after the issues encountered at the earlier stage were resolved.

The technology team provided a technical note that detailed the process involved in the migration. You can read the entire document by clicking here. The key takeaways being:

–    Understanding of data structure and account behaviour in legacy and new system is a key to successful migration.
–    It is advisable to always plan for mock drill before actual migration. This has to be accompanied by thorough testing and verification of data post migration.
–    Training should be given to users and they should be educated on account behaviour in the new system for accounts coming from the legacy system
–    Have patience.