
Ex- Machine Learning @ Inception AI
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
I am actively looking for my next job opportunity in machine learning research and engineering. My last job role was as a senior applied scientist at InceptionAI (a G42 company) where I contributed to designing and building multi-modal generative AI applications. I love building and analysing neural networks. I aspire to build secure and accessible artificial intelligence products and services.
University at Buffalo
Master’s Degree, Computer Science
January 1, 2013 – January 1, 2014
Jaypee Institute of Information Technology
Bachelor of Technology (B.Tech.), Information Technology
January 1, 2007 – January 1, 2011
Inception
Senior Applied Scientist
July 1, 2024 – June 1, 2025
Abu Dhabi, Abu Dhabi Emirate, United Arab Emirates · On-site
Hunch
Senior Data Scientist
March 1, 2023 – October 1, 2023
New Delhi, Delhi, India · On-site
Cactus Communications
Senior Data Scientist
February 1, 2022 – December 1, 2022
New Delhi, Delhi, India · Hybrid
Amazon
Applied Scientist II - Machine Learning
October 1, 2018 – January 1, 2021
Greater Boston
Amazon
Applied Scientist - Machine Learning
February 1, 2016 – September 1, 2018
Greater Boston
FactSet
Software Engineer II (Machine Learning)
September 1, 2015 – February 1, 2016
New York City · On-site
FactSet
Software Engineer (Machine Learning)
February 1, 2015 – September 1, 2015
New York City · On-site
Accenture
Software Engineer
December 1, 2012 – July 1, 2013
On-site
Accenture
Associate Software Engineer
July 1, 2011 – December 1, 2012
On-site
Recommender Systems: Big Data Matrix Factorization Using Apache Spark
December 1, 2021 – January 1, 2022
Technologies : Amazon EC2, Apache Spark, RecSys • Built a Recommender System on the Movie Lens dataset (20 million movie ratings) using Matrix Factorization. Performed matrix factorization using Apache Spark on Amazon EC2 cluster. • Evaluated the recommender systems model on Mean Squared Error.
A Transformer Chatbot Implementation
November 1, 2021 – December 1, 2021
Technologies : Tensorflow-2.0, Keras, Transformer • Built an Encoder-Decoder based chatbot using the Transformer architecture on the Cornell Movie Conversations dataset.
An Empirical Analysis of Sequence-to-Sequence Model Architectures for Machine Translation
October 1, 2021 – December 1, 2021
Technologies : Tensorflow-2.0, Keras, Bi-LSTM, Attention Mechanism, Transformer • Built different Seq-to-Seq based Encoder-Decoder modeling architectures with(out) attention mechanism, including the Transformer architecture for English to Spanish machine translation. • Analyzed performance using BLEU score.
An Empirical Analysis of different CNN Architectures for Multi-Class Image Classification
October 1, 2021 – December 1, 2021
• Built 7 different CNN architectures from scratch for multi-class classification task on the Fashion-MNIST dataset. • Analyzed the performance of each model architecture using confusion matrix.
Latent Attribute Inference of Twitter Users from Neighbours : Age Classification
January 1, 2017 – May 1, 2017
Technologies : Python, Support Vector Machines, Text Mining, Information Retrieval • Extracted Latent attributes of Twitter users using there tweet content and neighbourhood (followers and following). • Build features from the Latent attributes for classification of a Twitter user in the age category 18+ or 25+. • Used data of 360 Twitter users for training and validating the Machine Learning model. Collected data using Twitter API.
Simple Distributed Hash Table
April 1, 2014 – Present
Technologies : Java, Socket Programming, Android Application Development • Designed and implemented a distributed hash table using a simplified version of Chord Algorithm where each node is an instance of an Android Application. • Implemented ID space partitioning/re-partitioning, Ring-based routing, and Node joins.
Diabetic Level Prediction : Regression
April 1, 2014 – Present
Technologies: Matlab, Linear Regression • Designed and implemented a machine learning model based on linear regression to predict the diabetic level of a patient. • Features were based on physical, physiological and blood related measurements of a patient.
Totally and Causally Ordered Group Messenger with a Local Persistent Key-Value Storage
February 1, 2014 – Present
Technologies: Java, Socket Programming, Android Application Development, Python • Designed and implemented a Group messenger that preserves total and causal ordering of all messages. • Implemented vector timestamp at each process to maintain happened before relationships. • implement a key-value table that each device uses to individually store all messages on its local storage
Handwritten Digits Classification
January 1, 2014 – Present
Technologies: Matlab, Neural Networks, Support Vector Machines, Logistic Regression • Designed and implemented machine learning models based on Neural Networks, Support Vector Machines and Logistic Regression to classify hand written digits. • The handwritten digits were size normalized and centered in a fixed image of 28 × 28 size. • Compared the performance of the machine learning models based on various metrics.
Virtual Memory Module-PINT Operating Systems (PINTOS)
November 1, 2013 – Present
Technologies: C, Kernel Programming • Built the virtual memory module of PINTOS which overcomes the shortcomings of its small (4 MB) physical memory. • Implemented the page table using hash table to optimize the page lookup time to constant time. • Implemented the LRU clock algorithm for evicting frames improving efficiency by reducing page faults.
Question Answering System
November 1, 2013 – December 1, 2013
Technologies: Java, Apache-Solr, Stanford NLP POS Tagger, Java Wikipedia API • Created 3 indexes for people, places and films using Wikipedia info-boxes. The info-boxes are parsed using a SAX parser. Around 10,000 different Wikipedia documents are used to build the indexes and Solrj is used to index documents. • Used the Stanford NLP POS Tagger package to get the syntax of the question by the user and thus publish an answer according to user information need, type of question, solr index core etc.
Performance Analysis of Information Retrieval Models
October 1, 2013 – November 1, 2013
Technologies: Java, Apache-Lucene, Stanford NLP POS Tagger • Analyzing the performance of Vector Space and BM25 IR models on different metrics.
Multithreaded Web Server
September 1, 2013 – October 1, 2013
Technologies: C, POSIX Thread Programming, UNIX Socket Programming • Designing and implementing a multithreaded web server on the UNIX platform. • Implemented a queuing thread for incoming requests, scheduler thread to schedule the request and multiple worker threads to serve the request. • Implemented Shortest Job First and First Come First Serve CPU scheduling policies to serve the requests.
Wikipedia Indexer
August 1, 2013 – October 1, 2013
Technologies: Java, jUnit, SAX Parser, Regular Expression • Built a Wikipedia Indexer which can index Wikipedia articles under terms, author, category and links. • Performs extensive parsing on text such as removal of internal and external links, font styles, templates, section titles, lists using regular expressions. Also performs tokenization of words. • Retrieval can be done to obtain top-k documents. Supports Boolean retrieval and wild card queries.
Email Spam Classifier
March 1, 2012 – Present
Technologies: MATLAB programming, Support Vector Machines, Gaussian Kernel • Designed and implemented machine learning model using Support Vector Machine with Guassian Kernel to classify emails as spam or not. • The dataset is a subset of SpamAssassin Public Corpus. • Only body of the email was used to extract various features for the spam classification problem.
Machine Learning
Coursera
June 24, 2026 – Present
Algorithms: Design and Analysis, Part 1
Coursera
June 24, 2026 – Present
Cultural Fit Analysis
The candidate's project portfolio shows a strong inclination towards research and development in machine learning, with a focus on practical applications. The diversity of projects, from distributed systems to advanced ML models, suggests a curious and adaptable mindset. The experience in various companies, including startups and large enterprises, indicates an ability to adapt to different organizational cultures. The target role of ML Engineer aligns well with the candidate's demonstrated technical depth and project experience in building and deploying ML solutions.
Soft Skills & Operational Fit
The candidate's project descriptions indicate a proactive approach to problem-solving and a strong ability to design and implement complex systems. Experience in leading and contributing to multiple model releases suggests good operational fit and ability to work within a structured development lifecycle. Mentoring experience indicates leadership potential and knowledge sharing capabilities.