Software Engineer - Advance Analytics

RISHABH BANSAL

• Los Angeles, CA • 213-***-**** • ********@***.*** • LinkedIn

EDUCATION

Master of Science in Computer Science, 4.0/4.0 Jan 2020 - Dec 2021 University of Southern California, Los Angeles

Courses – Artificial Intelligence, Analysis of Algorithms, Web Technologies, Natural Language Processing, Deep Learning, Machine Learning Teaching Assistant – Analysis of Algorithms, Viterbi-USC Research Assistant – Center on Knowledge Graphs, ISI-USC Bachelor of Technology in Electronics and Communications, 3.8/4.0 Jun 2012 - May 2016 Guru Gobind Singh Indraprastha University, Delhi, India SKILLS & CERTIFICATIONS

• Languages Python, Java, SQL, PySpark

• Tools TensorFlow, PyTorch, Keras, Numpy, Pandas, Scikit-Learn, Matplotlib

• Big Data Spark, Dremio, Presto, PostgreSQL, MySQL, Hadoop, Hive

• Web Technologies HTML5, CSS3, Bootstrap, JavaScript, JQuery, AJAX, Angular, Node.js, Flask, REST API, JSON

• Miscellaneous Azure Databricks, Git, Linux(CentOS), GCP App Engine, AWS Lightsail, Android, Selenium, Adobe AEM 6.2

• Certifications Deep learning specialization (Coursera), ML (Coursera), Oracle Certified Java Programmer (OCJP 6) PROFESSIONAL EXPERIENCE

INFORMATION SCIENCES INSTITUTE, USC Los Angeles, CA Research Assistant Advisor: Prof Jose-Luis Ambite July 2020 - Present

− Developing Spark-Postgres mediator (data virtualization) to integrate & harmonize multi-source heterogenous data for learning.

− Creating a federated deep learning solution to process Schizophrenia neuroimaging data using TensorFlow.

− Researching query pushdowns and cost optimizations in mediators for federated learning.

− Technology – Python, Tensorflow, Java, Spark, Scala, SQL, Dremio, Presto, MySQL, PostgreSQL INFOSYS LIMITED Pune, India

Software Engineer – Advance Analytics Aug 2016 - Oct 2019

• Built scalable solutions for Fortune 50 CPG & Retail clients.

• Awarded top performer in Infosys foundation training in SQL, Python (97%) & Java full stack (94%).

• Geolocation Analytics: [Product: Infosys Geo-Tracker; Role: Senior Developer] May 2019 - Sep 2019

− Developed shipment tracking application to recognize & categorize shipment unloading points, halts and benchmark travel-times between consecutive geospheres.

− Designed scalable data pipeline to process terabytes of geospatial shipment tracking data.

− Technology – Python, PySpark, MySQL, SQL

• Data Quality Forecasting Project: [Client: Procter & Gamble; Role: Technical Lead] Jan 2018 - May 2019

− Led a team of 5; developed scripts in Python, PySpark, SQL to develop a scalable forecasting model & deployed it on Azure Databricks.

− Created time-series regression model to flag sales discrepancies. Identified sales driving feature that improved model accuracy by 6%.

− Implemented custom data partitioning in Spark to parallelly compute 500K+ models daily yielding performance increase of 85%.

− Decoupled training and prediction logic, optimized compute - resulting in savings of $350k+ in computational costs per annum.

− Awarded Infosys INSTA Award for exemplary performance across technologies in E2E project delivery.

− Technology – Python-ML, Sklearn, PySpark, Azure Databricks, SparkSQL, SQL, Hive etc.

• Image Analytics Project: [Client: Kroger; Role: Track Lead – Image Management] June 2017 - Dec 2017

− Created an automation soltuion in Python, Java to automate Kroger Image Management Track resulting in 6.5 FTE release.

− Developed OCR script to read product description; built a LSTM classifier for extracted description.

− Created image matching scripts to match & score product images using openCV2. Awarded Infosys Rising Star Award.

− Technology – Python-ML, OpenCV2, Tensorflow, Keras, Java, Selenium etc.

• Digital Learning Hub: [Client: Apple Inc; Role: Java Full Stack Developer] Aug 2016 - Apr 2017

− Developed services, servlets, templates, UI for Apple learning hub in Java Full-Stack with Adobe AEM 6.2 in Agile Scrum

− Created scripts for restructuring Apple learning hub to incorporate live copies for regional web pages.

− Technology –Java, JavaScript, JQuery, AEM 6.2, HTML5, CSS3, Bootstrap, AJAX etc. PROJECTS

ClipMe [Pytorch, Transformers]: Created an E2E application to auto-generate memes from unseen images based on CV & NLP. Solution pipeline incorporates four main modules – image to caption n/w, meme template classifier n/w, image caption to meme caption generator n/w, and a meme caption to audio mapper network. Blog GO Playing Agent [Python, RL]: Developed a 5*5 GO playing agent using reinforcement learning and tree pruning. Trained the agent for 70k games against random & greedy agents. Implemented minimax alpha-beta pruning to reduce branching factor. Achieved 98%-win rate. EBAY Web App [Angular, JS, HTML5, CSS3, Bootstrap, Node.js (Express), AJAX]: A responsive eBay search and catalogue web app hosted on GCP App Engine [App server: Apache; Web server: nginx; Frontend: Angular9 framework; Backend: Node.js]. App

Contact this candidate