A password will be e-mailed to you.

Creating new opportunities in ML and AI with Siddharth Das, Founder and CEO, Univ.AI

Siddharth Das, Founder and CEO, Univ.AI

Siddharth Das, Founder and CEO, Univ.AI

1. Tell us a little about the Univ.ai flagship Master ML & AI Program and the Data Science Leaders Program?

Our flagship part-time Master ML & AI program aims to prepare students and working professionals for the most challenging and rewarding careers in ML and AI. It is designed to closely mirror a graduate program in Data Science at a top US university.

Data Science Leaders Program is a first ever hire-then-train program in Data Science that we launched recently, to identify top talent early and then train them to become successful AI Engineers and DS Consultants with us. The driving force behind this program is to address our inability to find data scientists with backgrounds from other domains for our consulting practice. ​​As a part of this program, we assess applicants with backgrounds ranging from software development to biology for future promise as a Data Scientist. We make industry-leading job offers to them and then train them in our Master program on full scholarship. Upon completion of training they join Univ.AI’s consulting arm and begin work on state-of-the-art client projects as Machine Learning and Data Science consultants in India and overseas.

Based on the response for this program, we will likely extend this as a service for companies that wish to create top tier talent pipelines for the future.

For its first cohort, DSLP received over 5000 applications for the 10 positions we had on offer, from top universities and organisations across the country.

2. With so many courses in data science already on offer, what differentiates the programs at Univ.AI?

To illustrate, perhaps it’s appropriate to ask what distinguishes MIT from so many other tech institutions on the planet? We seek to make the caliber of education at MIT and Harvard accessible to deserving candidates globally at a fraction of the cost, especially those who are already working. We provide elite training with the world’s top professors, in a highly mentored environment. Everything at Univ.AI is live. Classes, labs, office hours, with real, live interactions with world-class instructors and mentors. Perhaps the closest representative description is that we are ‘a Harvard’ for ML and AI outside the walls of Harvard.

Most online programs lack this kind of orientation. Most lessons are through recorded videos, and once-a-week mentorship sessions shift the burden of learning entirely to the students. At Univ.AI, mentoring begins right at the Lectures. During lectures students work on TA-supervised problem solving sessions. Mentoring continues through Labs and office hours with Professors and TAs.

About 50% of the learning journey is completely mentored. Of the remaining, about 30% is peer-assisted and collaborative. Only 20% is unaided.

Mentored and peer driven learning as a % of the learning journey
Online programs – 10-20%
Universities – 30-40%
Univ.AI – 80%

Founding Faculty - Univ.AI

Founding Faculty – Univ.AI

3. How does Univ.AI plan to build its consulting arm and why was it set up?

Our consulting arm was set up to address the huge deficit in trained Data Science and AI talent worldwide. Our unique approach makes us perhaps one of the few such companies in the world that’s able to deploy highly trained Data Science and AI talent on client projects. We are able to help clients on boutique and specialized Data Science and AI problems as well as in large scale staff augmentation.

4. Elaborate on the skill gap that currently exists within the talent pool when it comes to technical data science skills?

Data Science and AI are deeply technical fields that require three kinds of expertise, Data Science, programming, and experience or expertise in a domain of concern. It’s therefore difficult for an employer to hire without considerable experience unless they are willing to invest into training, and then risk losing the resource as soon as they are trained.

University learning is still ramping up and availability of trained teachers suffer from the same volume constraints as that for other Data Science professionals. When the market for talent is as lucrative as it is for Data Science and AI – the average trained professional earns north of $300,000 in the US – it is difficult for all but the top universities to even offer high quality data science programs.

Working professionals suffer from an even harder time finding high quality learning opportunities.

5. Who are the founding faculty at Univ.AI?

Dr. Pavlos Protopas, Scientific Director, Institute for Applied Computational Science (IACS) at Harvard University. Pavlos also runs the Masters program in Data Science at Harvard.

Dr. Rahul Dave, formerly a lecturer at IACS, Harvard, now full time founder at Univ.AI

Dr. Raghu Meka, Associate Professor of Computer Science at UCLA. Raghu’s area of interest is in areas of theoretical computer science that are at the heart of Machine Learning and AI.

Dr. Achuta Kadambi, Assistant Professor of Computer Science at UCLA. Achuta is a rising star in computer vision, and runs the Visual Machines group at UCLA. He is also a cofounder of a Vinod-Khosla funded computer-vision startup Akasha.im

6. What motivated the leadership to launch Univ.AI?

Univ.AI was designed as an accessible and affordable alternative to top-tier universities, especially for working professionals. The founders of Univ.AI recognize that despite the proliferation of online learning programs that hope to capitalize on the enormous popularity of Data Science and AI, most fail to address the key learning issues that lead to high-quality outcomes.

Univ.AI seeks to address the central issues relevant to no-compromise, high-end learning in the field head-on, and seeks to meet cutting-edge benchmarks for outcomes.

7. What is the typical profile of applicants at Univ.AI?

Univ.AI doesn’t have a typical applicant. Applicants do have a strong technology bias, however they are drawn from diverse fields and different stages of learning, both among students and working professionals.

It is interesting to note that among our learners are professional data scientists from top companies, for example, Microsoft, and Adobe, who wish to add formalism and rigor to their training. At Univ.AI they can have the opportunity to do so without the high-cost or other constraints of enrolling into a top institution. Our entry barriers are considerable, since we do require our applicants to be well prepared. That said, we are fast, and we offer cohorts once a quarter, which makes an exceptional destination for such seekers.

8. Data science is relatively new and continuously evolving, how does one stay abreast with the latest developments in the field?

It’s also a very broad field. Staying abreast requires different things from different people. One might want to read blogs, be part of different online communities, and read research papers in their domain of interest. Those whose awareness needs are more high-level, can get by with popular articles available on blogs and other online resources. Being a part of active Data Science communities helps. Univ.AI is creating such a community as a part of its community learning initiative called the Geoffrey Hinton Fellowship – a learning, assessment and employment platform – inaugurated by none other than the godfather of AI, Geoff Hinton himself.

9. Which industries are ripe for disruption with data science and analytics and where do you see the field 5 years from now?

Well just about every field is being upended by Data Science and AI. So it’s hard to name one. But a good place to start is medicine, where from diagnosis, to imagining, treatment and drug discovery, the applications of AI will be huge. Security is another area which is already being heavily disrupted by computer vision. Customer care is being disrupted by NLP. Computer programming itself will probably be unrecognizable in a few years. Oil and gas exploration, crop yield management, climate forecasting, defect monitoring on assembly lines, process control in chemical plants, the applications of AI are endless.

The signature trait of where AI is likely to make a huge difference is where there’s lots of data available and where the relationship between input and impact are difficult to model using conventional means.

In 5 years, you’ll see a huge presence of AI in jobs that require human intelligence, which are simple, and resist automation. In terms of large scale presence of AI, customer care is one such area. Logistics is another. On the other end, you’ll also see AI able to do things that are very difficult or even hazardous. At its very best AI will free very skilled professionals from doing mundane things – like doctors writing prescriptions and treating coughs and colds – and instead being able to focus their talents where it is most deserved.



Send this to a friend