Project: Multicast ABR Controller (2020 - present)

Technologies/Languages: Typescript, Nodejs, java.

Platforms/Frameworks: Angular 10, Nestjs, Spring boot.

Designed and developed state of the art Angular UI for the controller part for MABR solution.

Designed and developed a Nestjs backend for the UI.

Designed and developed a RESTful backend for performing CRUD operations on network gateway related data.

Designed and developed RESTful backend for performing CRUD operations on video streams data.

Developed deployment artifacts - Dockerfiles and Kubernetes deployment files for deploying these services in kubernetes cluster

Developed an in-house redirection service which redirects ABR clients to correct Network gateway to save bandwidth.

Developed monitoring dashboards in Grafana to visualize network bandwidth savings and cache efficiency of Network gateways.

Designed and developed video manifest trimming service for live HLS and DASH manifests in java.

Project: Telemetry Data Pipeline

Developed spark batch jobs to aggregate user viewership data.

Spark jobs consisted of parsing and cleansing Avro and parquet data, aggregating the data and storing it into hive database.

Developed a reporting pipeline using Tableau. Pipeline consisted of reading hive data using Spark thrift server and doing incremental loads in Tableau.

Project: Manifest Delivery Controller

GSM REST API

Designed and developed NodeJS based RESTful application to

1. configure and manage backend services.

2. query backend microservices for statistics and exposing it external applications.

3. provide a unified login and authentication mechanism for REST API based on openid-connect, oauth2. Keycloak was used as the auth server.

Single sign on integration

Unified all third party apps such as grafana, kibana and in house UIs to use one single authentication service based on oauth2 - openid connect, where keycloak acted as the in house auth server.

Integrated ldap user federation into keycloak.

Added role mappings and privileges to manage access.

DevOps

1. Developed docker artifacts for multiple microservices.

2. Developed automation scripts to deploy docker based microservices into kubernetes clusters.

Project: Multicast Adaptive Bitrate (MABR)

Work: Video Streams Latency validation tool

Tool to test the end-to-end functionality and latencies of MABR.

Developed a NodeJS RESTful application to start multiple ABR (Adaptive Bitrate) HLS stream sessions.

Monitor the statistics and calculate time required for channel(video stream) to become popular(after number of views reached a certain threshold).

Monitor the time required for channel to be multicasted.

Developed an Angular 2 based dashboard to display the latencies, bandwidth usage without MABR and bandwidth usage with MABR

Technologies: NodeJS, Angular 2, Docker, MongoDB

Explain and help the students to complete their Data Structures assignments in JAVA.

Reviewing and grading the assignments.

Domain: Telecommunications

BI & Machine Learning:

Implemented Machine Learning algorithms for recommendation engines.

Algorithms Implemented: User-based collaborative filtering, Item based collaborative filtering, Logistic regression, TopN neighbours.

Technologies Used: R

Shell Scripting:

Developed Unix Shell Scripts (ksh93) to parse unstructured Ab-Initio text logs for extraction of performance related information of the ETL jobs thereby reducing the time required for analysis of logs by more than 50%.

Developed Automation and Deployment scripts for ETL tasks thereby reducing the efforts by 40%.

Developed Unix shell scripts framework for logging, reporting and monitoring tasks thereby reducing costs by 30%

Technologies : Unix Shell Scripting (ksh93)

Web Development:

Worked on development of a Analytics platform, designed for management level stakeholders of a financial website.

Technologies : Spring MVC, AngularJS, jQuery

Code Deployment and Releases:

Managed code releases and deployment to System Test environments.