Sanyam S.

Sanyam S.

BengaluruIndia

Machine Learning Developer

I am a deep learning expert in the fields of computer vision and natural language processing. I am an TOP_RATED (top 10%), full-time freelancer . I regularly create — and immensely enjoy creating — custom neural networks for my client's specialized problem domains! —————————— You are here because: —————————— - You have a small to mid-sized startup - Your startup is pushing the boundaries of what is possible, and you require deep learning techniques to make it happen - You have tried pre-built cloud ML/DL solutions and none work for your problem domain - You know that there is a massive gap between the type of products that AI titans (Google, Meta, OpenAI, etc.) have produced internally using their resources, and what is currently possible when applying those same products in real-world, practical AI applications in a specific domain - You need help in applying the latest deep learning techniques to your problem domain - You need a deep learning based minimum viable product (MVP) to attract investment money to your project ——————————— Sanyam is here because: ——————————— I love solving problems using deep learning! I spent many years in the field of scientific research, so I have 5 years of experience in the creative process of taking an idea from conception to production. Many of my clients have no experience in AI, find me before they have started their DL research efforts, and love how productive I can be from day one. I help my clients to: - Research cutting edge developments in the latest deep learning literature and open source projects - Choose the best DL framework and model architecture for their task - Curate custom datasets - Augment existing datasets - Create specialized DL models - Tune model performance - Create data processing pipelines - Build production quality neural networks I create value for my clients by being a one-stop-shop for coding and technical knowledge. For many of my clients I am their first technical hire. This means I architect and implement the initial technical infrastructure to set my clients up for future success. Once we have built an MVP and gained enough traction to obtain extra funding, additional technical resources are hired to speed product development. ————————— I use Pytorch, Tensorflow, and Keras to train my Deep Neural Networks, and analyze the data using Python, Numpy, P
Rating is 5 out of 5.
5.00 Jun 22, 2020 - Oct 21, 2020

"He did a lot of research on very difficult and newest cutting edge technologies. Kendryte K210 is a new AI chip for the edge. There is no existing knowledge on the field about AI on the edge on microcontrollers. Becouse its a new territory. He was able to push the hardware to the limits for some very complicated calculations and machine learning ( depth mapping and human detection). He adapted all existing machine learning and calculations and was able to fit it on this small and cheap microcontroller. He never gived up and always stayed positive! He will work for us further for the next version. And i hope for the next years to come! Excellent engineer!"

Private earnings
Rating is 1 out of 5.
1.00 Jul 30, 2020 - Aug 3, 2020

"When we entrusted him with this task, the timeline we had in mind was one week or week and a half to complete the 6 steps of the project. The first part of this project only contains a couple of functions. He have been logging hours in Upwork for 4 days, and you have practically not advanced from the first day. We do not know what he did on Saturday, or what he did today, since the code is nearly identical to that of Friday, and you have been logging the hours manually (around 12 hours). This project is trial and is very simple, and it does not include anything from neural networks or from adjusting models or anything. Also, he have uploaded the entire virtual environment to the Github repository, which is useless since the virtual environment of one computer does not work on the other. He just had to upload the configuration files. That shows that he doesn't have much experience with python virtual environments or with Github. We cannot be reviewing each line of the code to see if he forgot steps of the process, which happened, because that prevents us from doing our work. He can not even do the math himself, neither solve any problem alone. The idea when you hire someone is that you take away our work, not that he gives us more. After discussing about that, he blamed us, which is typical."

"There was no timeline or time limit from his side from day one. He kept on increasing the work and demanded more work outside of normally accepted and agreed by us from the start. 10 minutes calls went off to hours and chatting about his unclear requirements were throughout my day. I cannot talk throughout a day about unclear tasks. I completed the work on Friday itself. He insisted to advance with a way lot requirements which is not probable to be done in a day or so. He requires a consultor for his ideas first then go for a developer to complete them else its like a project without any head or tail."

Private earnings
Rating is 5 out of 5.
5.00 May 7, 2020 - Jun 1, 2020

"Nice cooperation"

Private earnings
Rating is 5 out of 5.
5.00 May 5, 2020 - May 12, 2020

"Sanyam delivered good work on this ML project. His communication was top-notch, he met all deadlines, and his skills were reasonably strong .Am happy with the work so will likely have additional jobs for him in the future."

Private earnings

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

Portfolio

Deep Wikipedia Search and Edit History modelling
Deep Wikipedia Search and Edit History modelling
Eye Tracking System
Eye Tracking System
Want to see more? Create Account
Sanyam S.

Sanyam S.

BengaluruIndia
9
Total Jobs
1,309
Total Hours

View profile

Machine Learning Developer

Specializes in
I am a deep learning expert in the fields of computer vision and natural language processing. I am an TOP_RATED (top 10%), full-time freelancer . I regularly create — and immensely enjoy creating — custom neural networks for my client's specialized problem domains! —————————— You are here because: —————————— - You have a small to mid-sized startup - Your startup is pushing the boundaries of what is possible, and you require deep learning techniques to make it happen - You have tried pre-built cloud ML/DL solutions and none work for your problem domain - You know that there is a massive gap between the type of products that AI titans (Google, Meta, OpenAI, etc.) have produced internally using their resources, and what is currently possible when applying those same products in real-world, practical AI applications in a specific domain - You need help in applying the latest deep learning techniques to your problem domain - You need a deep learning based minimum viable product (MVP) to attract investment money to your project ——————————— Sanyam is here because: ——————————— I love solving problems using deep learning! I spent many years in the field of scientific research, so I have 5 years of experience in the creative process of taking an idea from conception to production. Many of my clients have no experience in AI, find me before they have started their DL research efforts, and love how productive I can be from day one. I help my clients to: - Research cutting edge developments in the latest deep learning literature and open source projects - Choose the best DL framework and model architecture for their task - Curate custom datasets - Augment existing datasets - Create specialized DL models - Tune model performance - Create data processing pipelines - Build production quality neural networks I create value for my clients by being a one-stop-shop for coding and technical knowledge. For many of my clients I am their first technical hire. This means I architect and implement the initial technical infrastructure to set my clients up for future success. Once we have built an MVP and gained enough traction to obtain extra funding, additional technical resources are hired to speed product development. ————————— I use Pytorch, Tensorflow, and Keras to train my Deep Neural Networks, and analyze the data using Python, Numpy, P
Rating is 5 out of 5.
5.00 Jun 22, 2020 - Oct 21, 2020

"He did a lot of research on very difficult and newest cutting edge technologies. Kendryte K210 is a new AI chip for the edge. There is no existing knowledge on the field about AI on the edge on microcontrollers. Becouse its a new territory. He was able to push the hardware to the limits for some very complicated calculations and machine learning ( depth mapping and human detection). He adapted all existing machine learning and calculations and was able to fit it on this small and cheap microcontroller. He never gived up and always stayed positive! He will work for us further for the next version. And i hope for the next years to come! Excellent engineer!"

Private earnings
Rating is 1 out of 5.
1.00 Jul 30, 2020 - Aug 3, 2020

"When we entrusted him with this task, the timeline we had in mind was one week or week and a half to complete the 6 steps of the project. The first part of this project only contains a couple of functions. He have been logging hours in Upwork for 4 days, and you have practically not advanced from the first day. We do not know what he did on Saturday, or what he did today, since the code is nearly identical to that of Friday, and you have been logging the hours manually (around 12 hours). This project is trial and is very simple, and it does not include anything from neural networks or from adjusting models or anything. Also, he have uploaded the entire virtual environment to the Github repository, which is useless since the virtual environment of one computer does not work on the other. He just had to upload the configuration files. That shows that he doesn't have much experience with python virtual environments or with Github. We cannot be reviewing each line of the code to see if he forgot steps of the process, which happened, because that prevents us from doing our work. He can not even do the math himself, neither solve any problem alone. The idea when you hire someone is that you take away our work, not that he gives us more. After discussing about that, he blamed us, which is typical."

"There was no timeline or time limit from his side from day one. He kept on increasing the work and demanded more work outside of normally accepted and agreed by us from the start. 10 minutes calls went off to hours and chatting about his unclear requirements were throughout my day. I cannot talk throughout a day about unclear tasks. I completed the work on Friday itself. He insisted to advance with a way lot requirements which is not probable to be done in a day or so. He requires a consultor for his ideas first then go for a developer to complete them else its like a project without any head or tail."

Private earnings
Rating is 5 out of 5.
5.00 May 7, 2020 - Jun 1, 2020

"Nice cooperation"

Private earnings
Rating is 5 out of 5.
5.00 May 5, 2020 - May 12, 2020

"Sanyam delivered good work on this ML project. His communication was top-notch, he met all deadlines, and his skills were reasonably strong .Am happy with the work so will likely have additional jobs for him in the future."

Private earnings

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

Portfolio

Deep Wikipedia Search and Edit History modelling
Deep Wikipedia Search and Edit History modelling
Eye Tracking System
Eye Tracking System
Want to see more? Create Account
More than 30 hrs/week

< 24 hrs response time

Browse and buy related projects