Manish B. Email & Phone Number

Mountain View, California, Us

The world’s fastest Real Time Data Platform• Working on building Apache Spark connector for faster read & writes for Spark applications with Aerospike's Real Time Data Platform• Working on benchmarking graph traversal strategies for a Graph product built on top of Apache Tinkerpop and supported by Aerospike.• Built a Spark bulk loading process that allows customers to load graph data in Gremlin format using the Graph product into Aerospike DB.• Technologies: Scala, Java, Python, Spark, AWS, Docker, Gremlin

Palo Alto, California, Us

AppLovin turns mobile into a transformative advertising medium, using data gathered across screens to deliver a consistent experience tailored to individuals. The company's platform evolves as individual interests change – ensuring that our recommendations always align with what people really want. Every month, the company’s ad serving platform in combination with ML delivers recommendations to a billion consumers via dynamic ads that help brands keep pace with their customers. With AppLovin, advertisers can keep their brands relevant, drive value for consumers, and see meaningful results.AppLovin's leading marketing software provides developers a powerful, integrated set of solutions to grow their businesses. AppLovin enables developers to market, monetize, analyze and publish their apps. The company’s first-party content includes over 200+ popular, engaging apps and its technology brings that content to millions of users around the world. AppLovin is headquartered in Palo Alto, California with several offices globally.• Part of the DSP team that is responsible to build and deploy ML models at massive scale that are used to show relevant ads to customers on mobile apps.• The ads are kept relevant by tuning models based on consumer past interest such as payment prediction, install rate prediction, previous viewed ads, etc.• AppLovin is the ad partner layer between advertisers and bidders• On a daily scale, AppLovin handles over a billion ad impression via dynamic apps through its partner studios as well as through its DSP (Demand-Side Platform) layer that connects advertisers and publishers• Technical Stack: Python, PySpark 3.1, Spark ML, GCP, Weights & Biases, Java/Scala, API’s, Scikit learn, Keras, Tensorflow, Airflow, Jenkins, Dataproc, Grafana

San Jose, California, Us

* Retail Solutions Inc was acquired by IRI in October 2020.Retail Solutions Inc. is a Retail Data Science company focused on providing SaaS products to retailers and CPG vendors to transform high volume and diverse dataset into actionable insights.• Worked on building data engineering pipeline for predicting on-shelf availability of items in store thus improving sales and profitability of retailers and vendors.• The Spark driven pipeline runs Data Science Models built using Spark ML & outputs score for each product within a retailer dataset.• The score is used as a measure to predict availability of items in store & the future demand.• Focused on building Delta Lake components for fast ingestion and processing of data on a daily basis. The Ingestion job processes ~200M records for a single retailer vendor combination.• Worked on API microservices that integrate with Power BI and Tabular model to display metrics & key business indicators (KPI’s) that give a global perspective on the entirety of business & supply chain to retailers & vendors/suppliers.• Technical Stack: Java, Scala, Spark 2.4.4, Akka, Kafka, JSON, Java Spring Boot, Shell Scripting, Git, SQL Server, Jenkins, Azure, Docker, Kubernetes, Databricks on Azure, Jupyter notebooks.

Mclean, Va, Us

• Built the next-gen Fraud detection & Credit Scoring Microservice based architecture for credit card applications using Java/Scala, Spark and Akka.• The team is responsible for building a Microservice that enables fraud detection & credit scoring of applications in real time based on a score from Data Science models.• Responsible to build the Microservice layer/framework using Akka and Akka HTTP to score multiple applications in real-time.• Compute score by running each applicant record through Data Science models exposed as API's through docker containers.• Received multiple Spot Awards for ownership, delivery and customer focus.• Technical Stack: Java, Scala, Akka, Akka HTTP, Spark, Kafka, Shell Scripting, RDBMS, Jenkins, AWS, Docker, Kubernetes, ELK, Redis, Splunk, New Relic.

Mountain View, California, Us

• Research on Apache Hadoop 2.4 ecosystem & document necessary features necessary to migrate the hadoop services from ver 1.3 to 2.4• Wrote Hadoop services and API’s to migrate the Hadoop distribution.• Test the entire Orchestrator product for different Hadoop distributions like CDH4.5, HDP1.3 and Apache Hadoop 1.3, 2.4.• Technical Stack: Java, SQL, Spring, Maven, Shell Scripting, Git and Github, Jenkins, Hadoop, Map Reduce.

Conceptualized an Expertise Model for the Kaybus Product and submitted a proposal for the same. The Chief Architect accepted the proposal for integrating it with the company product.• Technical Stack: SQL, Hadoop, Ember.js, MongoDB, HTML5 & CSS3, Ruby, Ruby on Rails, Github, Jenkins.

Armonk, New York, Ny, Us

Technical Stack: Java, SQL, Shell scripting, IBM DB2

Master'S Degree, Computer Software Engineering

Engineer'S Degree, Computer Science

Machine Learning Nano Degree