Tech Adventures!

We have a wonderful post by our newest developer, (drumroll please) Devin Miller!  He put college on hold to come and work with NextDrop (and we are incredibly thrilled to have him here in Hubli, let me tell you).  Anyway, lets get to it. Here’s what Devin has to say about his experiences thus far.  

It has been a wild few months since I arrived here in Hubli at the beginning of September.  As I am still a university student, a year ago, I would have thought I’d be in school right now, continuing my studies.  I had no idea that I would be moving to India for 10 months to pursue working with the NextDrop team. 

 

Before arriving in India, I got to get my feet wet by exploring some of the NextDrop code.  I had a blueprint in my head of how the technology worked after reading through some of the blog posts and the business plan.  Crowd-sourcing, water, social-impact… it all sounded so exciting.  I was informed I would be part of the scaling team here at NextDrop.  Toyota’s “Lean manufacturing” came to mind by reducing waste and overproduction; these ideals played a huge role in the work I am doing here.

 

So what do I do here at NextDrop?  Thej, our lead programmer, and I, make up the tech team and work together to maintain and to build the next version of the NextDrop technology.  Upon arrival here, there were many concerns about our system and more importantly, the data that we were collecting. 

 

Where do all of our messages go? Do they arrive on time? Do the residents even open our messages?

 

These were just some of the questions that were posed.  We could provide a lot of answers after generating delivery reports from our voice calls and from our SMS gateway.  We took all this data on messages/calls and created a program to help us process all that data, turning the data into useful information.  This information told us how many messages were delivered, the status of the message, and whether they were delivered on time or not.  We were able to reduce the number of invalid messages being sent and also better inform our customers to be more receptive to our messages.  When I first arrived here at NextDrop, about 75% of our messages were delivered successfully. 

 

Only 75%? What about the other 25%!?

 

Invalid phone numbers, mismatched SMS template being rejected by the NDNC registry, duplicate messages accidentally being sent, just to name a few.  We were able to decrease the number of void messages being sent by addressing these reasons and still work towards improving the success rate of these messages today.  Our message delivery success rate is now at a steady 90%.  

 

Collecting the data and the availability of the reports within the company, greatly improved our customer service resolution time.  We were able to provide the customer with information that we did not have previously, for example, a customer may ask why they never receive our messages; from the delivery report, we could report back to them that their message inbox is full and that’s why they don’t receive any messages from us.  (This has happened!)  More importantly, we could report back to the customer in a timely manner.

 

This is just one example of collecting data and performing analysis, thus turning it into information to make data-driven decisions.  The coming back Lean principle, reducing waste, and cutting costs, are important for any company, especially a start-up.

 

A lot of my time is spent programming and maintaining the administrative dashboard that is the command and control center for the NextDrop backend.  One of the most important things I have learned is that the user interface (UI) is essential!  It is everything!  If you want to develop a successful product that users want to use, the UI is of the utmost importance.  We have been able to make drastic changes in how we interact with our dashboard, and now we have come to rely on it.

 

Our database a few months ago looked different than it does today.  Although it is still incomplete, we are slowly starting to fill in all the gaps.  With no central location for your data, such as a database, it is impossible to concatenate enough data and make sensible and logical decisions.  That was the case when I first arrived here.  We did not have enough information about where our residents lived exactly to bill efficiently.   There was a huge lag time in between finding an area to bill, actually contacting/finding the customer, logging the information in the field, and entering the data into a computer.  (This was probably the biggest bottleneck in the process.  Do not allow data entry to pill up!  The sooner you have the data entered, the more equipped you are to make better decisions.)

 

Changes to the administrative dashboard and entering all of the data into one central location drastically reduced this lag time.  We were able to bill more efficiently, by collecting information from residents, such as landmarks and neighborhoods (sub-sections of a valve area).  We also performed SQL database queries to find areas receiving good service and that have not been billed recently, meaning they are ready to be billed.  Recently, we have set records on how many customers we have billed in a single day.  Go Billing Team!

 

If there is one thing to take away from this, it is this:

The more data one collects (in a centralized location) and turns this data into information, the better data-driven decision-making becomes.

 

I know our blog may seem quiet at times, but I can assure you, we are here and we are working hard.  So much to program . . . so little time . . .

 

 

Devin