The post covid world requires new forms of communication - with the motivation to bridge the gaps between people, families and communities around the world. It has impacted all sectors with biggest impact on education, healthcare, business and social media interactions.
There are several apps which have been developed and out in the market. each one of these apps suffer from common challenges of scalability, performance and most of all - it needs all users to be on the same system.
take case of education - teachers and students need a computer enabled system and should be on the same application. most of these applications are costly and hence limits the access to a larger section of the society around the world.
Worldwide, 50% of students don’t have a computer at home. In sub-Saharan African, 89% of learners don’t have a computer at home. Source: UNESCO International Teacher Task Force, April 2020
Our motivation is to build an open-accessible setup which can maximize the audience reach of audio-visual content and provide a cost effective solution to deliver maximum impact across all geographies.
What it does
Dialoge empowers every individual, teacher, student and other professionals to create content in real-time and engage with a wider audience across social media, text, email etc.
We can start a live stream at the click of a button by simply going to the browser on the desktop or mobile phone, and then share the real-time streaming youtube link across to the audience - hence reaching an unmatched number of audience in minimum possible time.
The added benefit of streaming live is that the audio-visual content is saved for on-demand viewing also. e.g. a teacher works with a student for a follow-up session on difficult topics. the same session is made available on-demand so other students can also learn from the same. this saves valuable time for the teachers and also continues to build a knowledge base.
We have used youtube for setting this up, however, the same could be done with a complimentary aws services.
How we built it
Our primary aim was to move away from depending on complex software or customised SDKs which require high level of customisations and development effort, cost of initial development, licensing etc, as well as running costs.
We used open source technologies - nodejs, reactjs, ffmpeg and aws infrastructure to build this.
Challenges we ran into
setup of ffmpeg was challenging and we had earlier tried to work with heroku and azure and the service kept failing. we were almost giving up hope, however, we read about the EC2 Graviton and with reference to the GitHub page, where it specifically noted about ffmpeg performance improvement. hence we decided to make another attempt to bring this project to life.
Accomplishments that we're proud of
learning about graviton, setup of servers and infrastructure, getting ffmpeg to work in stable environment
What we learned
ec2 graviton can be key to our solution. we did some basic cost estimate using aws estimator and the service is approx 5-10x less than zoom.us subscription
What's next for Dialoge
build a consumer ready application for web and mobile, which we can showcase with focus groups in education, social and business sectors for pilots and feedback.