Nihal Poosa

● Developed backend services using Spring Boot in Java 21, integrating PostgreSQL for data storage, and extensively using Protocol Buffers in both persistence and messaging. Used Jackson for JSON serialization and deserialization. ● Implemented cluster auto-scaling on Amazon EKS Managed Node Groups and built a remote stateless daemon leveraging STOMP WebSockets and also HTTP for robust server communication.

● Lead the team of Spike Analysis and supervised juniors. Spike Analysis attributes the sudden spikes in website visit data of a brand to the ads watched of the brand before a spike occurred. Spark, Elasticsearch and NodeJS were used. ● Worked on a Movie & TV Show search, which supported popularity, typing mistakes and searches as you type using Elasticsearch (a real-time search engine). Built the Web APIs in NodeJS. It was primarily used by AOSP-based-TVOS. ● Worked on a differentiating and competitive feature, Creative Resonance, of our Ad Insights product, which tells how much a user is resonating with an ad by measuring how regularly they watched the ad. Used Elasticsearch to store the data and NodeJS for the APIs. Extensively used Apache Spark to do the analytics on the Ad viewership Big Data. ● Suggested a visual way to split the Creative Resonance, by network or show watched during the ad, daypart, etc., to find the networks, shows or dayparts causing higher resonance, which was put in production. ● Contributed to the TV-Guide team, by building Web APIs in NodeJS which served Program Airings, TV Stations, etc., data of any dates in seconds from an Elasticsearch database of 2 years’ worth data.

● Developed a web page which can render a 3D model reconstructed from a series of DICOM images in stereoscopic split- screen view for mobile VR, using Three.js, a WebGL JavaScript library, to create and light the scene, and load the model. ● Implemented controls for a Bluetooth connected joystick, to rotate and zoom in, using Gamepad API of HTML5.

● Assisted in developing a Facebook Messenger bot, an automated consumer support through which the user can shop fashionable clothes and shoes by chatting with it. Built the web-hooks in a Django server. ● Used api.ai to do the NLP like getting the attributes of a product such as color, brand, size from the message sent by user. ● Developed the conversational structure of the fashion shopping between a consumer and the bot, which asks all the choices they would be interested in. Also, implemented a feature to assist the consumer in style tips.