Neel Deshpande

Boston, MA 646-***-**** adczvg@r.postjobfree.com linkedin.com/in/neel-deshpande/ Available: May – Dec 2020

EDUCATION

Northeastern University, Boston, MA Sept 2019 - Present Khoury College of Computer Sciences Expected Graduation: May 2021 Master of Science in Computer Science, GPA: 3.83

Relevant Courses: Programming Design Paradigm, Algorithms, Information Retrieval, Natural Language Processing

Savitribai Phule Pune University, Pune, India May 2018 Bachelor of Engineering in Computer Science, GPA: 3.62 TECHNICAL KNOWLEDGE

Languages: C, C++, C#, Java, Groovy, Javascript, Python Databases/Frameworks: mySQL, PostGRESQL, MongoDB, SpringBoot Web Technologies: JQuery, ReactJS, MarionetteJS, BackboneJS, GoJS, HTML, CSS Certifications: Java Core/Advanced, Python, Machine Learning, Neural Networks and Deep Learning WORK EXPERIENCE

eQ Technologic Inc., Pune, India

Software Development Engineer Oct 2018 - Jun 2019

• Implemented redesign of eQubeMI’s UI from legacy stack to new stack employing MarionetteJS and BackboneJS

• Developed a data migration activity in Java to read data from a JSON datasource and write it in eQ’s format (MLV)

• Created REST APIs for fetching details about transactions on the eQubeMI server Rammer.ai, presently Symbl.ai, Pune, India Jul 2018 - Sept 2018 Data Science Intern

• Built word-level sequence-to-sequence English language generating RNN using LSTM units for speech to text translation under the direction of the Chief Data Scientist, achieved average F1 score of 0.409 for the pilot model

• Evaluated performance of translation model with Google’s Word2Vec word embeddings using cosine similarity and locality sensitive hashing techniques

• Performed data preprocessing using Pandas and NLTK for a model that adds punctuation to artificially generated English PROJECTS

3D Infinite Runner Game with Self-Adaptive Game Artificial Intelligence Sept 2017 - May 2018

• Developed simple infinite runner game in Unity 3D using C# that adapts to individualistic gaming style of player and generates obstacles accordingly

• Employed Dynamic Difficulty Adjustment and Weighted Random Forest algorithms to tune the Reinforcement Learning model governing physics of generated obstacles

Creche Management System Smart India Hackathon, 2018

• Built web portal in a team of 6 using ReactJS for parents to look up nearby creches and remotely monitor their children under the Rajiv Gandhi National Creche Scheme of the Ministry of Women and Child Development of the Indian Government

• Exploited Google’s Object Detection API with TensorFlow to fine-tune a learning model to recognize babies in images

• Listed as an Award Winning Solution by the Indian Government’s Ministry of Human Resource Development at https://mic.gov.in/micoutcome.php