I'm a CS Engineer, a Python Geek, and a Machine Learning enthusiast. I love playing with algorithms and data structures. You can check my spoj profile here. I have been to ACM-ICPC Asia Regionals twice (in 2014 and 2015 respectively) and each time my team was among the top 50 from about 300+ teams. I have published two research papers you can check them in my publications section.
Ankush Bhatia
604, Block-F, Accurate Wind Chimes
Narsingi, Hyderabad, Telangana.
Phone : +91-8000290624
E-mail : ankushbhatia02@gmail.com
MS in Computer Science • Pursuing
Current Grad Student at Georgia Tech. Specialization : Machine Learning.
B.Tech in Computer Engineering • May 2017
I completed my undergraduate education from Charotar University of Science and Technology, Gujarat.
Machine Learning Engineer• April 2019 - Present
I'm a part of the Machine Learning Group at Qualcomm. I'm working on the dev part of the AndroidNN HAL for Qualcomm devices.
Machine Learning Developer• September 2017 - March 2019
I worked mainly in Document Structuring, Information Retrieval and Image Processing. My major work involved working in core Image Processing (OCR, Noise Removal, Deskew), structuring readable/scanned PDFs (Financial Documents) into dataframes such that they're easily extractable. Used basic NLP and classification techniques to extract useful information from financials. Also worked briefly on NER tagging for financials and a basic search engine to query over a large dataset of documents. Technologies used : Basic Python Machine Learning libraries like Scikit-learn, Scipy, etc. Pandas and Numpy for handling and structuring data, elasticsearch for indexing and storing, Tesseract for OCR, and other basic nlp modules like NLTK, gensim, etc.
Machine Learning Developer• June 2017 - September 2017
I made Chatbots for healthcare. My work here mostly involved : NLP, Signal Processing, Data Engineering, Django, AWS. I have made several Machine Learning APIs like Speaker Recognition, Text Summarization, Text Chunking and Named Entity Recognition, Bayesian networks for Diagnosis, etc. ML Models and algorithms on which I have worked and working on are : One Class SVMs, GMMs, Bayesian Networks, HMM, POS Tagging, Text Chunking, Summarization using K-means and TextRank, etc.
Python Developer(Intern)• January 2017 - May 2017
Making Chatbots for healthcare. My work here mostly involved : NLP, Signal Processing, Data Engineering, Django, AWS. My work as an intern involved less ML work and more of DevOps where I had to deploy django codes on ElasticBeanstalk, work in EC2 and Lambda, etc. Main ML which I used during my internship was Text Summarization, Signal Processing, POS Tagging and Chunking for Named Entity Recognition. ML Models and algorithms : POS Tagging, Text Chunking, One Class SVMs, Text Summarization using TextRank.
Python Developer(Intern)• May 2016 - July 2016
Developed Recommender System for the startup. Technologies : Python, Scikit-learn, NetworkX, NLTK. Used Community Detection to overcome the coldstart problem. Used Neural Networks for semi-supervised learning i.e for updating user's movements in the netowrk space. ML models and algorithms on which I worked on here were : Graph Clustering, Regression, Neural Networks.
Part of the Machine Learning Team
I worked in the Machine Learning team of the project where we majorly worked on Information retrieval and NLP.
Technologies Used : Python 2.7, Sklearn, Scipy, Numpy, AWS EC2, Pandas, Tesseract, Google Vision API, Flask.
Designed the entire Backend of the Bot
Technologies Used : Python 2.7, Sklearn, Scipy, Numpy, AWS EC2, Pandas, Django.
Developed Backend of the bot
Technologies Used : Python 2.7, Sklearn, Boto3, AWS, Django, DRF, Scipy, Numpy.
Developed Backend of the bot
Technologies Used : Python 2.7, Sklearn, Boto3, AWS, Django, DRF, Scipy, Numpy.
Developed ML Backend of the app
Technologies Used : Python 3.5, Sklearn, Scipy, Numpy, NetworkX, IGraph.
My Github Profile
Check Out Some of my open source projects on my GitHub profile.Click Here
My Spoj Profile
I was into sport programming for 2 years of my undergrad studies. You can check my SPOJ (Sphere Online Judge) profile here. Click Here