Aashay Sachdeva

A system and method for matching players in multiplayer online games, accommodating asymmetric entry fees and payouts, varied geographies, time zones, and currencies, while employing skill-based matchmaking for fair play.

An intelligent system and method for matching players in multiplayer online games, utilizing player statistical parameters, historical data, and demographic information. The system dynamically adjusts matchmaking objectives to create challenging and engaging games.

An article discussing the opportunities and challenges in developing multilingual and multimodal language-vision models tailored for Indian languages and contexts.

An exploration of using multi-armed bandits in online systems, highlighting their ease of understanding, implementation, and ability to explore a wider range of options.

Platforms, systems, devices, software, and methods for processing and analyzing real-time data to predict near-future events using ensemble machine learning, combining models with varying aggression levels for enhanced timeliness and accuracy.

A guide on causal inference using quasi-experiments, emphasizing their importance when traditional A/B testing is not feasible.

A guide on essential metrics for tracking and evaluating Data Science projects.

Details the development and launch of Cutouts on Product Hunt, showcasing the implementation of machine learning directly within web browsers.

Discusses how the Data Science team at Kubric focused on smart automation to significantly reduce the time designers spent on grunt work.

A tutorial on using the Tensorflow Object Detection API for creating custom object detection/segmentation models and deploying them.

Explores variational dropout as a method to estimate uncertainties in neural network predictions and investigate the existence of subnetworks.