• Productionized bot performance analyzer ML model (MACS) to measure 100% of 1st and 3rd party bots, achieving industry-leading adoption and driving optimal outcomes, • Revamped Real-Time Visitor Monitoring Platform from VM-based to Kubernetes, cutting computing resource usage• Orchestrated real-time calculation of visitor scores using data science models applied to sessions, maximizing campaign goals through ML-driven consumer invites.• Dynamically managed chat invite percentages based on agent availability, optimizing consumer engagement and minimizing queue wait times• Implemented enhanced observability measures on existing services, reducing incident clutter by 60% and enabling proactive resolution of potential issues.• Knowledge Leveraged : Kubernetes, Pachyderm, Seldon, Scala, Python, Kafka, Cassandra, ElasticSearch, Prometheus, Grafana, Google Cloud, Hadoop, Spark, Apache Beam, Airflow, Locust, Hazelcast

- Update Django 1.6 to 2.2 LTS and python2.7 to Python3.7- Implement SOA for multiple variation of requirements- Consult testers to improve test plans by automation using postman and selenium- Coach junior developers - Constantly implement new features keeping standard Software Development Industry Standards. - Tech Tags: python, django, mysql, redis, docker, jenkins, nodejs, system design, scrum, postman, selenium, git, yodiz, docker-swarm

- Using Tornado server to handle up to 8k concurrent requests, having 4M of requests everyday.- The system is distributed, scalable and fault tolerant- Implement caching mechanism to fasten up reporting - Separate components in microservices using Docker- Introduce codereview, code coverage with TDD approach which becomes a prerequisite of CI- Refactor legacy codes with complexity optimization, maintaining OOD principles and proved design pattern- Initiate and maintain bug database, take part in RCA, collaborate with customer to support.

- Backend API development with Nodejs and deployment in AWS.- The challenge was to run video transcoding using FFMPEG to convert video files and- serving them efficiently using cloudfront and AWS S3 using algorithmic techniques and tweaks.

Responsible for polishing and adding new features in the Newscred CMS, a content management system build on Django framework with other cutting edge technologies both in front end (Websocket.io, Angularjs) and back end (Nodejs, RabbitMQ, Celery, Redis) with Memcache object caching and Varnish reverse proxy. that gives clients to create custom news portal based on licensed and non fulltext contents.