Nishant S.

Nishant S.

BengaluruIndia
Job Success
Top Rated

Deep Learning Expert | Full Stack Computer Scientist

🏆 Expert-Vetted, Upwork Top 1% ! ▶ Nishant Sinha: A senior hands-on engineer and scientist with 20 years of experience building a variety of deep learning, web/distributed systems at scale, taking products to the end user / market and publishing cutting edge research. I finished my Ph.D. from Carnegie Mellon and worked at IBM Research, building AI solutions and distributed systems tools. I founded OffNote Labs to help businesses with cutting-edge, deep technology solutions. I’ve spent most of the last decade building deep learning systems (across vision, text and speech domains), formulating data science problems and shipping solutions for clients across the globe. I bring a very unique combination of skills to the table: ↪ I’m a senior AI engineer: Extensive experience around model building (Transformers, GPT3, ViT, ASR, CLIP, Diffusion), compression/optimization (Pytorch, TensorRT, Deepspeed, ONNX) to Docker/Kubernetes/AWS/GCP. Python, JavaScript, now Go and Rust. Strategic problem solving with cutting edge deep learning and blockchain technology. ↪ As an Architect - I go from ambiguous specs to a clean, modular design of a complex system. Plan a system as a whole before optimizing the parts. Keep things simple. ↪ As a Researcher - I can distill out consumable stuff from the most math dense, cutting-edge academic text. I can explain and write about complex ideas and systems pretty well. Published internationally over 20 years. Solving complex, multi-disciplinary problems is my core strength. ↪ As a Product Entrepreneur - I deeply care about the product strategy: who buys it, who consumes/uses it. I hustle to bridge the gap between cutting-edge tech and impactful products. I spot, lead and mentor deep tech talents. --- My projects include a mixture of creative, algorithmic problem solving and building across the tech stack. Recent projects involve: GPT3, Natural Language Generation from structured objects, ChatGPT, Prompt Chaining to decompose complex tasks, Prompt engineering to match text styling. Stable Diffusion - img2img. ✔️ Please have a look at the Deep Learning section of my profile to learn more about my specific skills and past engagements. Since a year or so I’ve been working on blockchain related projects — Layer 2 solutions, threshold cryptography, MPC, privacy-preserving computation. ◈ I’ll be delighted to talk with you: - if you are eager to bring revolutionary products and cutting-edge tech to market - if you have an ambitious project, requires non-standard skill set and expertise. - if you think I’m the best candidate to get your task done. Look forward to hearing about your next amazing project. --- "Nishant is incredible to work with. He is a fount of machine learning wisdom and knowhow. His involvement with our project was instrumental in taking it to new heights." "Nishant is everything I was hoping for: (1) truly a skilled, experienced expert on AI and (2) truly a kind, thoughtful person to work with. Thanks so much for the great work."
Rating is 4.6 out of 5.
4.60 Dec 18, 2017 - Jan 30, 2018

"Nishant had a great impact of the success of the feature we requested! Thank you!"

Private earnings
Rating is 5 out of 5.
5.00 Jun 21, 2017 - Sep 5, 2017

"Excellent resource. I have hired 15-20 contract employees through Upwork over this past year and Nishant has proven to be a value to my operations over and over again. It is difficult sometimes to find trustworthy, experienced and skilled workers, especially when they can be halfway around the world. Nishant has been able to effectively communicate his ideas, tasks, recommendations and explain difficult concepts in lay man's terms. Job well done, I am hiring him again for another project."

Private earnings
Rating is 5 out of 5.
5.00 Jan 3, 2017 - Apr 18, 2017

"Incredibly hard working and knowledgeable. We rehire!"

Private earnings
Rating is 5 out of 5.
5.00 Mar 16, 2017 - Mar 20, 2017

"Will work again in the future"

Private earnings
Rating is 5 out of 5.
5.00 Mar 11, 2017 - Mar 12, 2017

"Most excellent. Will engage again very soon."

Private earnings
Dec 6, 2015 - Feb 15, 2017

No feedback given

Private earnings
Rating is 4 out of 5.
4.00 Apr 21, 2016 - May 11, 2016

"All was good, thx"

Private earnings

Nishant S. has more jobs. Create an account to review them

Portfolio

Build a neural network based query analysis and prediction engine. Writeup: https://medium.com/@ekshakhs/has-deep-learning-solved-search-yet-fb6490abfd1 Given a natural language query, maps it to the top 10 product categories in the catalog. Use these categories to do cascaded product search. Based on combined semantic (word vectors, RNNs) and syntactic search (BM25, tf-idf). A single neural model works across product verticals and categories. Optimized to reduce memory and latency. Build using PyTorch. Deployed on AWS.
Neural e-commerce search
Using Tensorflow based Transformer architecture for English to French translation. Extend the default beam search engine to generate multiple, diverse translations.
Improve neural beam search for diverse translations
Want to see more? Create Account

Skills

Nishant S.

Nishant S.

BengaluruIndia
Job Success
Top Rated
37
Total Jobs
3,554
Total Hours

View profile

Deep Learning Expert | Full Stack Computer Scientist

Specializes in
🏆 Expert-Vetted, Upwork Top 1% ! ▶ Nishant Sinha: A senior hands-on engineer and scientist with 20 years of experience building a variety of deep learning, web/distributed systems at scale, taking products to the end user / market and publishing cutting edge research. I finished my Ph.D. from Carnegie Mellon and worked at IBM Research, building AI solutions and distributed systems tools. I founded OffNote Labs to help businesses with cutting-edge, deep technology solutions. I’ve spent most of the last decade building deep learning systems (across vision, text and speech domains), formulating data science problems and shipping solutions for clients across the globe. I bring a very unique combination of skills to the table: ↪ I’m a senior AI engineer: Extensive experience around model building (Transformers, GPT3, ViT, ASR, CLIP, Diffusion), compression/optimization (Pytorch, TensorRT, Deepspeed, ONNX) to Docker/Kubernetes/AWS/GCP. Python, JavaScript, now Go and Rust. Strategic problem solving with cutting edge deep learning and blockchain technology. ↪ As an Architect - I go from ambiguous specs to a clean, modular design of a complex system. Plan a system as a whole before optimizing the parts. Keep things simple. ↪ As a Researcher - I can distill out consumable stuff from the most math dense, cutting-edge academic text. I can explain and write about complex ideas and systems pretty well. Published internationally over 20 years. Solving complex, multi-disciplinary problems is my core strength. ↪ As a Product Entrepreneur - I deeply care about the product strategy: who buys it, who consumes/uses it. I hustle to bridge the gap between cutting-edge tech and impactful products. I spot, lead and mentor deep tech talents. --- My projects include a mixture of creative, algorithmic problem solving and building across the tech stack. Recent projects involve: GPT3, Natural Language Generation from structured objects, ChatGPT, Prompt Chaining to decompose complex tasks, Prompt engineering to match text styling. Stable Diffusion - img2img. ✔️ Please have a look at the Deep Learning section of my profile to learn more about my specific skills and past engagements. Since a year or so I’ve been working on blockchain related projects — Layer 2 solutions, threshold cryptography, MPC, privacy-preserving computation. ◈ I’ll be delighted to talk with you: - if you are eager to bring revolutionary products and cutting-edge tech to market - if you have an ambitious project, requires non-standard skill set and expertise. - if you think I’m the best candidate to get your task done. Look forward to hearing about your next amazing project. --- "Nishant is incredible to work with. He is a fount of machine learning wisdom and knowhow. His involvement with our project was instrumental in taking it to new heights." "Nishant is everything I was hoping for: (1) truly a skilled, experienced expert on AI and (2) truly a kind, thoughtful person to work with. Thanks so much for the great work."
Rating is 4.6 out of 5.
4.60 Dec 18, 2017 - Jan 30, 2018

"Nishant had a great impact of the success of the feature we requested! Thank you!"

Private earnings
Rating is 5 out of 5.
5.00 Jun 21, 2017 - Sep 5, 2017

"Excellent resource. I have hired 15-20 contract employees through Upwork over this past year and Nishant has proven to be a value to my operations over and over again. It is difficult sometimes to find trustworthy, experienced and skilled workers, especially when they can be halfway around the world. Nishant has been able to effectively communicate his ideas, tasks, recommendations and explain difficult concepts in lay man's terms. Job well done, I am hiring him again for another project."

Private earnings
Rating is 5 out of 5.
5.00 Jan 3, 2017 - Apr 18, 2017

"Incredibly hard working and knowledgeable. We rehire!"

Private earnings
Rating is 5 out of 5.
5.00 Mar 16, 2017 - Mar 20, 2017

"Will work again in the future"

Private earnings
Rating is 5 out of 5.
5.00 Mar 11, 2017 - Mar 12, 2017

"Most excellent. Will engage again very soon."

Private earnings
Dec 6, 2015 - Feb 15, 2017

No feedback given

Private earnings
Rating is 4 out of 5.
4.00 Apr 21, 2016 - May 11, 2016

"All was good, thx"

Private earnings

Nishant S. has more jobs. Create an account to review them

Portfolio

Build a neural network based query analysis and prediction engine. Writeup: https://medium.com/@ekshakhs/has-deep-learning-solved-search-yet-fb6490abfd1 Given a natural language query, maps it to the top 10 product categories in the catalog. Use these categories to do cascaded product search. Based on combined semantic (word vectors, RNNs) and syntactic search (BM25, tf-idf). A single neural model works across product verticals and categories. Optimized to reduce memory and latency. Build using PyTorch. Deployed on AWS.
Neural e-commerce search
Using Tensorflow based Transformer architecture for English to French translation. Extend the default beam search engine to generate multiple, diverse translations.
Improve neural beam search for diverse translations
Want to see more? Create Account

Skills

As Needed - Open to Offers

< 24 hrs response time