Rudradeb Mitra, is the Founder of Omdena a global platform where a community of changemakers building AI solutions to real-world problems through collaboration.
What was your initial inspiration for launching Omdena?
There were three main reasons:
To use AI for social good and build solutions for the people at the bottom of the pyramid, to give opportunities to AI enthusiast no matter where they live to work on real-world projects and to show that collaboration is way more powerful than competition and is the best way to build inclusive, ethical and adoptable AI solutions.
The ultimate goal of Omdena is to build the future of work and education – to create a world where no matter where one lives, she has equal access to work and opportunities.
The goal of Omdena is for enthusiasts, who have never met each other, and who are located in different parts of the world to come together, collaborate and solve big societal problems. Can you describe the logistics of having all these people working together?
The key to making collaboration work is to select the right people – we go through a 2-stage selection process based on various criteria. Once the right people are selected, the rest is quite easy – provide them the right infrastructure, tools, and processes to collaborate and let them self-organize. As described by one of Omdena’s collaborators his experience, ‘What seems chaos at first, one starts to appreciate the experience and realize the relevance of working in a self-organizing environment, where crucial conclusions and optimal solutions eventually emerge as winners. It is a very well welcomed and valuable experience. It makes me proud to be a part of it.’- Vjeko Hofman, Omdena Challenge participant from Croatia.
How do you decide which challenges to tackle?
We first select the right organization to work with. The criteria for selecting an organization is that it is tackling a big global problem, has the ability to make a change at the ground level, and is able to use the AI/ML model that we build.
The details of the challenge is decided through an iterative process, where we look at various criteria to check if it makes sense to solve the problem using AI/ML. The availability of data is not a mandatory condition, as part of the challenge, the community tries to find the data. One of the collaborators, Anastasis Stamatis from Greece, wrote ‘The data is there. Incomplete, unstructured at times, but it is there, waiting to be found. All we need to make an impact, then, are the right people in the right place’.
One previous challenge was deciding where to install renewable power systems in Nigeria, this would help over 100 million people without access to stable power. What type of data did you look at and what were some of the insights that you gained?
We tried to identify areas in Nigeria that are most vulnerable and need immediate solar power. To do so we divided Nigeria into clusters (based on census data) and identified areas that are dark at night, more than 15km away from the national grid is densely (or reasonably) populated and has good sun energy. We built an online model which can be found here.
Thanking the work done, Ademola Eric Adewumi founder of the organization part of the Nigeria challenge wrote ‘Omdena’s AI Community Changed Our Lives’.
One can read more details and all the articles about the work here.
India has one of the worst track records when it comes to sexual harassment against women. Omdena was a partner in finding safe routes for women to take in major Indian cities. Could you talk to us about the type of data and machine learning that was used to generate these safe routes?
We partnered with an organization called Safecity India who has a mobile app, that men and women already use to report sexual harassment. Till now they have over 15000 reported incidents of sexual harassment – the data contains the location, time, kind of incident and free text to share the experience. We used the data to generate heat maps of incidents and then used a route planning algorithm to reach the nearest safe area (hospital, police station, or metro station).
According to the Safecity team; ‘Omdena’s community achieved in two months what we tried for two years’.
One can read more details and all the articles about the work here
Omdena’s most recent challenge is detecting wildfires in the Amazon with AI. This was an 8-week long machine learning project that pulled together a team of 47 data scientists from 22 countries. What were some of the major accomplishments that your team were able to achieve?
We worked with a Brazilian company, Sintecsys, fighting forest fires. In this case, they had access to a large number of data. We were able to build models (95% accurate on validation set) that could detect smoke. When asked about his experience working with Omdena, Osmar, Head of Innovation of the Brazilian company Sintecsys said, ‘Outstanding in many ways. Omdena is, from now on, the official AI partner of Sintecsy’.
One can read more details and all the articles about the work here
What are the qualifications someone should have before they approach your organization to help?
The main qualification we look for in people is an eagerness to make this world a better place. They can either belong to an organization or can be an individual.
For someone who does not want to join Omdena, but wishes to start a similar project in their local community, do you have any recommendations?
My only recommendation is just to start. I have met so many people who wish to start or want to do something but never end up doing. So, if one wants to do it, then do it 🙂 Also, one can reach out to me if you need any help or guidance.
In 2018, you published a book called “Creating Value with Artificial Intelligence: Lessons Learned from 10 yrs of Building AI Products and Overcoming Data, Adoption, and Engineering Challenges”. Could you tell us more about this book and what you discuss in it?
I wrote this book to show the power of AI and how it can solve social problems. This book bridges the gap between the commoner’s world and the AI technical world. It is for anyone who wants to dig deeper into the practical aspects and explore the value of AI and build intelligent products for real-world use cases. I talk about:
– Where can AI or intelligent machines add Value?
– What real-world problems can be solved and what problems are harder to solve using existing AI algorithms,
– A brief explanation of the most applicable algorithms and for which use cases they are most suitable,
– How to engineer and architect AI systems,
– How to overcome some of the major challenges with gathering the data, developing the product, and making users adopt the product.
The above steps are explained through use cases taken from banking, insurance, energy, sales, healthcare, and other sectors. Almost all the knowledge and use cases shared in this book are based primarily on my personal experiences.
I am also glad that the book got a lot of great reviews (19/20 reviewers gave it 5 stars in Amazon), some of the ones that I like:
“This book can transform your vision on how you can even imagine building or make smarter things.”
“This is the book of the year from my side. Thank you”
“A great book, about a subject that is not easily grasped by the vast majority of people, in a very light way, accessible to everyone.”
Any Final words?
I am super excited about the future and looking forward to seeing where Omdena takes us.
Authors Note: I was thrilled to be able to profile Rudradeb Mitra and the work behind Omdena, an important project that is changing the world in so many positive ways. To learn more visit Omdena.
Also take a peek at “Creating Value with Artificial Intelligence: Lessons Learned from 10 yrs of Building AI Products and Overcoming Data, Adoption, and Engineering Challenges” for a different type of book on AI.
- Lior Hakim, Co-founder & CTO of Hour One – Interview Series
- The Smart Enterprise: Making Generative AI Enterprise-Ready
- Flick Review: The Best Instagram Hashtag Tool to Boost Reach
- U.S. Imposes Export Restrictions on NVIDIA Chips to Certain Middle East Countries
- Tanguy Chau, Co-Founder & CEO of Paxton AI – Interview Series