Mukundhan Srinivasan

Engineering curiosity, code & creativity. Fedrer Fan. WIP Sri Vaishnavan.

profile_img.jpg


I’m back at NVIDIA in the US for a second innings. Currently, I lead the Generative AI Software business for NVIDIA in North and Latin America.

Before relocating to the United States, I worked in Toronto as a Senior Data Scientist, specializing in machine learning at datacenter scale. My interests are broadly in machine learning methodologies and system engineering. My past research pursuits included language generation, specifically in dialogue systems for generation and evaluation, as well as concepts of generalization. I have a deep fascination with language. See publications

As the Lead ML Architect at Paytm Labs in Canada, I spearheaded advanced machine learning projects. Before this, as a Senior Architect at NVIDIA, Bangalore, I collaborated with a team of researchers and engineers, developing deep learning models, platforms, and pipelines optimized for GPU-based training and inferencing.

In the early stages of my career, I co-founded Market Learning Labs, building developer tools for computer vision and NLP applications. Despite its eventual failure due to non-viable market fit, this experience was invaluable. My career began as a Research Assistant at the Indian Institute of Science, Bangalore, where I worked on OCR and TTS systems for Indic languages.

For over a decade, I have been an engaged IEEE volunteer, dedicating my efforts to the scientific community. I previously held the role of Vice-Chair for the Membership Development Committee in IEEE Region 7 (Canada). In 2020, I was elevated to the status of Senior Member (SMIEEE).

Beyond my professional focus in CS/ML, I am deeply interested in the structure of economies, Bitcoin, the history of societies, the human condition, and South-Indian classical music. Some of my favorite musicians: Maharajapuram Santhanam, E. Gayathri, U.Srinivas, Mohammed Rafi, and SPB. I dedicate a significant amount of time to horology and the appreciation of fine writing instruments.

I am devoted to studying and practicing Sanatana Dharma and Vishistadvaitam as propounded by Bhashyakarar and Srimad Vedanta Desika under the Rahasya-traya guruparamparai.

After Tendulkar and Dravid’s retirement, my interest in cricket waned, although I occasionally catch a test match. I’m a fervent admirer of Federer, particularly his unmatched inside-out forehand and point/play orchestration. I continue to learn and play chess (mainly because it is another tab in the browser) and recreational tennis.

I used to live in (then beautiful) authentic part of Bangalore closer to CTR’s benne dose and Veena store’s idli chutney before moving to North America. As a Southerner, my fondness for an aromatic cup of filter kaapi (yes, Kaapi!) is profound, and I’m easily persuaded by a good one.

My social media profiles are not a reflection of my current activities or interests due to muted enthusiasm for frequent posts and limited time. I maintain this space for all practical purposes. Of course, MOAMO.

You can reach me at new_string AT immsrini DOT com.

def remove_last_three(s):
    return s[:-3]

original_string = "mukundhan"
new_string = remove_last_three(original_string)
print(new_string) #replace output to get username for email address

Just a heads up: I’m currently not looking for new job opportunities. This inbox is for personal correspondences only



Past Research Colaborators: Prof. Mitesh Khapara and Prof. Balaraman Ravindran, IIT Madras; Prof. Venkatesh Babu, IISc Bangalore; Prof. Vinay P Namboodiri, IIT Kanpur; Prof. Debdoot Sheet, IIT Kharagpur; Dr. Tavpritesh Sethi, Stanford Medical; Prof. Anurag Agrawal, IGIB; Dr. Vidur Mahajan, CARING Research


news

Oct 22, 2023 I now lead NVIDIA’s Generative AI SW busienss for North & Latin America. Back to building. :rocket:
Mar 16, 2020 After a short stint at Paytm Labs, I’m back at NVIDIA!
May 07, 2019 I’ve joined Paytm Labs, Canada to work on cool ML problems in the AdTech, SaaS, ecommerce and payments space.
Feb 25, 2019 Our paper titeled Efficient Video Classification Using Fewer Frames has been accepted at CVPR 2019 (25.2% acceptance rate) for oral (< 6% of accepted papers). This is work done with Shwetha and Mitesh at IIT Madras. Paper here