Reflections from Bangladesh
Notes from an insightful and productive trip visiting our Apps Development Team
I got back back home last week after what was a wonderful trip to Bangladesh visiting our Apps Development Team and our extended Left family there. Fortunately, I’m no stranger to this beautiful country — this was my 6th visit to Bangladesh in four years, and on every visit this place tends to grow on me more.
Our Apps Team in Bangladesh are mostly Android developers and are our ‘internal users’ of the RightMesh platform. As internal users, they do a lot of the alpha testing of the platform and developer library and provide valuable data on how we need to refine our Android libraries and developer tools to make them more user friendly. This trip was beneficial in gathering those insights.
The main objectives of this trip were to review the app prototypes the team has been working on with the RightMesh platform over the past few months, to bring the team up to speed on some of the recent developments of the project, and to realign on our mutual goals for the short term. The last several months have been extremely hectic for all of us as we complete our Token Generating Event (we’re at the last round, almost there!), but it was time to reset and reflect on what the most important work should be at this juncture in our project development.
We had a large contingency visiting Bangladesh this time around. From Vancouver, we had Dr. Jason Ernst (RightMesh CTO and Chief Scientist), Ben Hughes (Mesh Engineer), Marinna Lachapelle (Admin & Culture Assistant) and myself (Chief Product Officer). Dr. Lucien Loiseau (Lead Mesh Networking Scientist) also joined us from Singapore. The first week was largely spent meeting our Apps teams in both Dhaka and Khulna, exploring ideas and insights about the RightMesh project, and reviewing all the prototype apps they have developed so far.
RightMesh Updates
To kick things off, I provided an update on the project and the roadmap, then Jason reiterated the vision of RightMesh in a presentation to the entire office, and we concluded with a detailed Q&A whiteboard session for the RightMesh team. There were lots of good questions about how micro payment channels work, the Superpeer, and what the evolution of the platform looks like.
Lucien gave a great talk on Delay Tolerant Networks (DTNs) including what they are, the use cases that are most relevant for the store-and-forward type scenarios for mesh networks, and plans of introducing DTN-like APIs to the RightMesh platform.
App Prototypes Review
It was very helpful (and a lot of fun!) to review the app prototypes the team has developed over the past few months including: several multiplayer games that could be played by users without any internet connection, an app to directly exchange digital business cards, a decentralized voting app, a Pocket-like app that saves offline web content that could then be shared that with others directly, and an app that can stream music across a mesh network.
The team has been using an alpha version of the RightMesh library to develop these prototypes, and with the platform still in nascent state, the team did face a few obstacles during development. Despite the challenges, they were able to develop some really good functioning prototypes, and consequently from those reviews we took back a lot of good insights for the platform — such as the need to include new functionality in the library to make debugging easier and the need for a different set of APIs that can send larger sizes of data across a mesh network, but in a less reliable fashion when it is not time-critical.
Development Challenge
Our second week in Bangladesh was spent working closely with the team to determine our mutual goals for the next several months ahead. While the prototypes developed to this point have been largely experimental on the RightMesh platform, the focus now is towards building prototypes that can be tested in-country and that can gather important technical and user data required to develop the platform.
We believe hackathons are one of the best ways to discover developer feedback on the platform, and so, as part of a one-day development challenge, the team took on an exercise to build a simple RightMesh system comprising of three different components:
- A RightMesh node acting as a mobile sensor device that would collect data [client]
- A RightMesh node with data connectivity acting as a collector device [internet selling node]
- And a server component that would collect data and process it into permanent storage [app Superpeer]
Sensor devices would pass on data to the collector device periodically whenever they were in close proximity with each other, and the collector would send this data up to the Superpeer for processing. The most important criteria: links between the sensor and collector, and the collector and Superpeer would be unpredictable and inconsistent, and so the apps developed would need to handle connection failures.
Seven groups with 3 team members each worked on the exercise for a day (and they all did a really great job!). The following day, we held a team-wide retrospective on the challenges faced and best practices in developing RightMesh apps which turned out to be a really productive session. Ben came up with a simple yet important set of guidelines for building RightMesh prototypes:
- Prototype Early: Generate better ideas for your solutions.
- Iterate: Learn from your prototypes to plan your next steps.
- Simplify: Avoid doing work that the RightMesh library can do for you.
- Avoid Flooding the Network: Optimize the amount of information you send between peers.
- Be Resilient: Remember that states and links in the mesh can change at any time.
Focus Groups
I always find on the ground user research fascinating, and fortunately I was able to conduct a few focus group discussions around internet and mobile data usage with students from universities in Khulna. I had done similar surveys back in 2015, and quite a few things have changed since then — notably that many more students are accessing their internet through fixed connections available in their dorms (which they set up and pay for on their own).
Students pay a fixed monthly fee for unlimited usage on a 1–5 Mbps connection, which is horribly slow. Mobile data connections have much higher speeds, but students tend to avoid them as much as possible because of the much higher cost; as as a result, 90% of the students’ internet consumption is from their slow, fixed internet connections. Cost is still more important than speed, and there is no incentive to use internet on a mobile device. We gathered lots of fresh foundational data to analyze and utilize as we refine our token model.
Next Steps
Over the next several weeks, we will be working with the Bangladesh team to determine the next set of prototypes and experiments to work on that will not only be useful for testing developing on the platform, but will also be beneficial to test and learn specifically from their home market. And, we all agreed the one thing we need to ensure in all of this is to be agile and stay lean, building in units and learning on the way.
One of the biggest benefits we can leverage from our team in Bangladesh is our ability to learn and test building RightMesh networks directly on the ground in one of our early target markets. As is the case with many crypto projects, there is much uncertainty and many more questions than answers.
But, we intend to navigate through this uncertainty one incremental step at a time, testing and retesting networks models until we determine the optimum platform-market and token-utility fit.
Overall, it was an amazing and insightful visit, and it was great to meet our Apps teams both in Dhaka and Khulna once again! Not only did we have some really deep and productive discussions about the project, we also had lots of fun and good times on the way. As one of our core mantras, we believe in working hard and playing hard together!
I’m really excited for the next few months ahead working closely with this dedicated team as we incrementally move forward in achieving our mission with the RightMesh project. Looking forward to being back in a few months, and until then, পরে দেখা হবে! (See you later!)
Note: Many thanks to our colleagues Rakib Islam, Ayesha Siddika and the rest of the Bangladesh team for hosting us — we are blown away every time with your kindness and your warm hospitality.
