Amol Kelkar, Developer in Redmond, WA, United States
Amol is available for hire
Hire Amol

Amol Kelkar

Verified Expert  in Engineering

Deep Neural Networks Developer

Location
Redmond, WA, United States
Toptal Member Since
July 31, 2019

Amol is a seasoned technologist with 18 years of experience. He has worked in artificial intelligence, as a computational neuroscience research scientist, as a machine learning engineer, CTO, entrepreneur, vice president of engineering, and product strategist. He has invented eight patents and written two publications in his field. Amol has also founded, and participated in several startups.

Portfolio

Got It AI
Deep Neural Networks, GPT, Generative Pre-trained Transformers (GPT)...
i3ai.org
Amazon Web Services (AWS), Pandas, Google Cloud, PyTorch, Python
Office Space
Amazon Web Services (AWS), React, JavaScript, PostgreSQL, Redis, Python...

Experience

Availability

Part-time

Preferred Environment

Jupyter, Git, PyTorch, Python

The most amazing...

...machine learning algorithm I've developed learns disentangled representations using brain-like laterally connected neurons.

Work Experience

Head of AI R&D

2018 - PRESENT
Got It AI
  • Led AI R&D efforts at Got It AI, a conversational AI company that offers an award-winning, state-of-the-art platform for building fluid, conversational task-oriented virtual agents.
  • Envisioned, designed, and built Progressive Flow Discovery, the technology behind Got It AI's products for automated conversational flow discovery from historical conversational logs. This system uses large language models (LLMs) like Bert.
  • Envisioned, designed, and built Article Bot, a fluid conversational interface over a set of articles/knowledge bases. This technology uses generative deep learning models.
  • Led a team of ML engineers, product engineers, and applied ML scientists.
Technologies: Deep Neural Networks, Generative Pre-trained Transformers (GPT), GPT, Natural Language Processing (NLP), Artificial Intelligence (AI), Python 3, PyTorch, Clustering, System Architecture

Founder and AI Research Scientist

2017 - PRESENT
i3ai.org
  • Founded a state-of-the-art machine learning research organization. He is the lead researcher.
  • Utilized novel methods for reinforcement learning and representation learning.
Technologies: Amazon Web Services (AWS), Pandas, Google Cloud, PyTorch, Python

CTO

2013 - 2017
Office Space
  • Managed software development, web infrastructure, and product.
  • Built a sophisticated ETL infrastructure to synthesize data from various semi-structured data sources with my team. The web application provides a listings discovery and communication platform used by more than 100,000 users each month.
  • Grew the Office Space listing count by ten times, and leads by 12 times the previous amount.
Technologies: Amazon Web Services (AWS), React, JavaScript, PostgreSQL, Redis, Python, Ruby on Rails (RoR), Ruby

Director of Technology

2010 - 2011
King of the Web
  • Ran the technology, engineering, and web operations.
  • Developed seven releases from prototype to public beta in less than five months.
  • Hired and managed an excellent engineering team.
  • Contributed to product definition and strategy.
Technologies: Redis, PostgreSQL, Ruby on Rails (RoR), Ruby

Founder and CEO

2008 - 2010
Aflatune
  • Oversaw recruiting and managing technical and marketing resources, business development, and investor relations. Aflatune was a music streaming service with development offices in the US and India.
  • Aflatune used state-of-the-art machine learning techniques for music recommendations and scalable architecture for the streaming platform.
Technologies: Machine Learning, jQuery, JavaScript, CSS, PostgreSQL, Ruby on Rails (RoR), Ruby

Software Engineer

2000 - 2008
Microsoft, Corp.
  • Developed products and services such as Internet Explorer, NetDocs, InfoPath, Office, and Windows Live (Bing).
  • Invented and developed eight patents.
  • Developed software using .NET and C#.
  • Interviewed and mentored software engineers.
Technologies: C#

OfficeSpace.com

http://officespace.com
I developed the OfficeSpace.com platform. I managed software development, web infrastructure, and product. My team built a sophisticated ETL infrastructure to synthesize data from various semi-structured data sources and a web application that provides listings discovery and a communication platform used by over 100,000 users each month.

Languages

Ruby, Python, Python 3, C#, CSS, JavaScript

Frameworks

Ruby on Rails (RoR), Ruby on Rails 4

Libraries/APIs

PyTorch, Pandas, jQuery, React

Other

Machine Learning, Deep Reinforcement Learning, Natural Language Processing (NLP), Image Processing, Deep Neural Networks, Artificial Intelligence (AI), Clustering, System Architecture, GPT, Generative Pre-trained Transformers (GPT)

Tools

Git, Jupyter

Platforms

Amazon Web Services (AWS)

Storage

Redis, Google Cloud, PostgreSQL

1999 - 2001

Master's Degree in Computer Science, Computational Neuroscience

University of Texas at Austin - Austin, Texas, USA

MARCH 2018 - PRESENT

Deep learning by Nando de Freitas

Oxford

NOVEMBER 2017 - PRESENT

Reinforcement Learning by David Silver

DeepMind