MK & his team at Code Z are currently building DataGenie, a Gartner recognized Augmented Analytics product, that combines Machine Learning, Generative AI & Automation to autonomously mine actionable business insights from vast data.DataGenie was recognized in the 2021 & 2022 Gartner® Market Guide for Augmented Analytics Tools and other Gartner research materials. DataGenie is being used by large enterprise businesses including Logitech, Farfetch & LexisNexis.Checkout https://datagenie.ai for more details.

Built an end to end ML platform for Zalando that supports multiple stages of the ML life cycle including data discovery, exploratory data analysis, data pre-processing, experimentation with with multiple ML frameworks & algorithms, hyper parameter tuning, A/B testing, batch & API inference. Also implemented multiple high value ML use cases for Zalando including campaign product recommendation, classification models to determine propensity for customer loyalty programs, product pricing forecasts, customer care traffic forecasts

Solution Architecture for Ericsson's Big Data Analytics & Artificial Intelligence practice - with leading Telcos & Cable Operators in North America.Conceive, Architect & Evolve Ericsson's Lean Lambda Reference Architecture for Big Data Analytics & AI - leveraging state of the art open source technologies including Spark, Kafka, Nifi, Cassandra, Elasticsearch, H2O, Tensorflow.Technical Supervision for Ericsson Expert Analytics product deployment for T-Mobile. Technical Supervision for Ericsson's consulting engagements with Tier1 operators - including Machine Learning based Alarm Correlation & RCA, Reverse Logistics etc.

Architecting and Delivering Big Data Analytics solutions for Ericsson customers in Region North America including T-Mobile, Comcast & Rogers.

Senior Specialist in Ericsson responsible for BSS/OSS Open Source Technologies adoption. I'm a Open Source evangelist and believe in building reliable platforms for Ericsson around robust open source technologiesChief Architect of Ericsson's Big Data Analytics Product which includes a home grown Business Logic Definition layer integrated with Hadoop, HBase and a home grown OLAP Platform based on Pentaho Mondrian & Apache FlexArchitect for a generic, container agnostic O&M (FCAPS) platform for Ericsson, built with Apache Zookeeper at the core.Also have used Apache Camel in the Big Data Platform for extensible data-ingestion and data-exposureWorked with Apache Storm for streaming data ingestion, prototyped a meta-data driven Storm based real-time aggregation frameworkBuilt a meta-data driven, Subscriber Profiler framework based on MapR HadoopArchitecting introduction of HBase/MaprDB in the Ericsson Analytics product to replace usage of a proprietary RDBMS.Represented Ericsson in MapR Customer Advisory Board.One of the main architects of Ericsson Expert Analytics, a market leadingTelco Analytics Product.Technical supervision of large teams of expert software developers & architects.

Architected an end-to-end Big Data Analytics platform for Wifi Analytics for Comcast in AWS. We built the following in a span on one year:- Poller infrastructure to regularly poll statistics from millions of Wifi Access Points- Data ingestion layer based on Gobblin to on-board raw data from multiple Comcast data centers into AWS - also converts data on the fly from various native structures to Avro- A thin but powerful configuration driven abstraction layer above Apache Spark - with connectors to read/write data from S3, HDFS, ElasticSearch, Cassandra, processing based on SQLs. The idea was to utilize the power of SparkSQL and give back the power of Big Data Processing to the huge community of SQL experts - almost all of the Wifi Analytics data processing (joins, correlations, aggregations) can be done in Spark by anyone that knows just SQL- ElasticSearch + Kibana based exploratory analytics for the internal teams to explore raw data, identify new trends, new thresholds and data patterns - this served as a Data Playground and helped the internal teams define the analytics use cases- Used both Cassandra & ElasticSearch for the aggregated data store - Cassandra for the usecases with well defined read patterns & ElasticSearch for the ones with OLAP like read patterns- Django REST API framework to expose data out of ElasticSearch & Cassandra- D3/Node.js based Visualization framework

One of the main architects for the home grown NMS (Network Management Systems) platform for Midas. We used open source netsnmp stack & built a generic multi-device management platform which was successfully used to manage DECT based Wireless in Local Loop systems, Fiber in Local Loop systems and DSLAMs