![Profile](https://weekday-user-pics.s3.us-east-2.amazonaws.com/profile-images/default.jpeg)
Alok Kothari
AI/ML, Crypto
10.2
Years of Experience
Education
carnegie mellon university
Companies
posted, south park commons, on deck, aiblockchain consulting, harmony, apple, carnegie mellon university, adelement, qatar computing research institute, university of stuttgart
Reach out to Alok Kothari via Email, InMail and SMS drip
by installing Chrome extension
Alok's contact details
Email (Verified)
kotXXXXXXXXXXXXXXXXXom
Experience
2022 - 2023
posted
Founder
LLM-based (Generative AI) Copilot for $2T Revenue Cycle Management Workflows (medical billing) industry.
2021 - 2023
south park commons
Member
Founder Fellow. Cohort #2
2020 - 2021
on deck
Fellow
Cohort #7
2020 - Present
aiblockchain consulting
Founder
Built a "RAG" Retrieval Augmented Generation that they deployed for building customer support chatbots for their clients. More LLM/RAG projects that cannot yet be disclosed.
2018 - 2019
harmony
Co-Founder
building, hiring and everything else required to make Harmony successful.
2015 - 2018
apple
Machine Learning Engineer (Siri)
* Formulating and researching models for natural language understanding using state of the art machine learning (deep learning) techniques. * Building machine learning models for production * Work on all parts of model deployment pipelines (data, train, test, deployment) * Mentoring machine learning interns. (2 directly supervised projects were among those selected for presentations to highest level of management chain)
2013 - 2015
carnegie mellon university
Graduate Research Assistant
There are multiple coreference resolution metrics for measuring the quality of a coreference resolution system. We independently defined a yardstick to assess quality of the output of a system. This yardstick was validated with a human study. Performance of each metric v.s. quality of output was evaluated against this yardstick, and recommendations about quality of metrics themselves were made.
2013 - 2013
adelement
Software Analyst (Machine Learning)
Modeled and implemented hotel recommendations for a major online travel site.
2012 - 2012
qatar computing research institute
Research Associate
Worked on extracting comment-tweets from twitter: tweets which are comments on specific news articles by twitter users. Such comments add diversity of opinion to comment section and drive user engagement. System has been adopted by Al Jazeera.
2011 - 2012
university of stuttgart
Research Assistant
Worked on using unsupervised word sense disambiguation to improve information retrieval. e.g. knowing whether 'bank' means financial institution or a river bank can improve ranking of documents in a search engine. Worked on models to do such a disambiguation automatically and in a unsupervised manner. First project was to validate improvement in information retrieval (i.e. ranking of documents), if at the query level, non-compositional phrases are identified. e.g. 'hot dog' is non-compositional phrase which is different from a 'dog' that is 'hot'. We identified and demonstrated that identification of such phrases improves document rankings.
Experience
17 Skills
Algorithms
Big Data
Bonds
Derivatives
Founder
Information Retrieval
Java
LaTeX
Linux
Machine Learning
Matlab
Natural Language Processing
NLTK
Python
Statistical Modeling
Statistics
Trading
Education
2013 - 2015
carnegie mellon university
Master's degree
Computer Science (Language Technologies Institute)