Shashank Kumar Mishra Email & Phone Number

Bengaluru, Ka, In

Bengaluru, Ka, In

Bengaluru, Karnataka, India

Working in the lending domain to enhance the efficiency of the loan approval process for customers.

Bengaluru, Karnataka, India

API Design and System Optimization: Led the design and implementation of APIs for various machine learning features, significantly improving backend performance by optimizing PostgreSQL configurations, implementing caching strategies, and connection pooling, reducing response times from 2.5 seconds to 1.2 seconds.Enhanced Security and Deployment: Implemented robust rate limiting, authorization, and authentication mechanisms to enhance system security, and streamlined deployment cycles from weeks to days through API automation.

Bengaluru, Karnataka, India

● Collectibles with quest(to increase retention period) improvements.To raise awareness about collectibles as rewards in quest since only ~25% DAU is aware of collectibles reward in quest,increased retention period by 30%. ● To increase quest collectible usage since only 50% of quest winners are using collectible rewards. Improve awareness of skip quest task feature Currently, only 13% of the Quest Participants are aware of the skip task feature and of them 45 -50% convert in spending Gems to skip a task. To ensure that the user stays on the quest popup and highlighting the skip button, I have worked on making users aware of the quest skip task feature which increased the revenue to 23% of existing.

Delhi, India

Worked on microservices to enhance the scalability of a chat application for a large number of users by using WebSockets for bidirectional communication, resolving causal dependencies, and implementing database-level caching for frequent requests. Additionally, created an API function to extract names from URLs and fetch chats based on specific parameters, and customized a WordPress theme for better user engagement with default posts, comments, and feed links.

Delhi, India

● I worked on computer vision to improve the efficiency of the detection of medical instruments.I used yolo(image detection algorithm) to detect the medical instrument which has improved the accuracy of the model from 80 to 95 percent.These systems were deployed in real world to help the doctors to save time during operation. ● I have worked on a recommendation system based on previous searches in the search history.Algorithm like BERT and transformer were used to show the nearest recommendation to the searches made by the user in real time.

Bengaluru, Karnataka, India

● The objective behind the project was to forecast the money spent on R&D of the company . The features used were revenue generated by the company for previous years,total headcount and industry name embeddings. The models that have been used were Random Forest and Linear Regression.The predicted amount was a bit larger than the previous amount of money spent on R&D. ● The objective behind the project was to automate the tagging of the news. Earlier the news were tagged manually. Gradually the data set was becoming very large so it was difficult to tag it manually, so Machine Learning came into the picture to automate it.The preprocessing steps were done like removing stop words and html tags and other special characters. Different features were created out of the data set.The models used were advanced deep learning which works over the preprocessed data.

I was preparing for GATE Examination and focussing on Computer fundamentals