Past Meetup

Technology Choices for Data Science Products

This Meetup is past

73 people went

Location image of event venue

Details

About the Speaker : Ashutosh Prasar
Ashutosh is a senior Data Engineer working at DBS bank. He is a masters in Computer science from NUS, Singapore and has working experience in Computer Network, Telco and financial domain.
Working as a Data Engineer, he has worked extensively on Hadoop, Spark, Solr and other Big Data technologies. His interest areas are building and tuning scaleable data products. He loves playing/watching Cricket and running.

Synopsis:
Exposing big data as APIs can be quite challenging as concurrent queries need to be responded in sub-second to keep the user interested. The selection of correct noSql DB is quite critical to the this. I will talk about some of the noSql database categories and how to select one for your purpose. I will also discuss some of the concepts of a time series database like Druid.

About the Speaker : Dr Emin Aksehirli
Dr. Emin Akşehirli is working as a full stack data wrangler at DataSpark. He is preparing data, developing algorithms, productionize them using DevOps techniques, and support them. He is a data scientist, an experienced Scala and Java developer, and a former Java trainer.

Synopsis
In DataSpark, we have to experiment and work with large amounts of data on HDFS. The performances and the workflows of the data science tools that has became the de facto standards, such as jupyter and R, are not always enough. We have been using Apache Zeppelin notebooks to do data science at scale for a while. I will be talking about how we leveraged Zeppelin to integrate our JVM and Spark-based product with our exploratory processes.

Purpose
There are ample tutorials and seminars publicly available that teach the specific data science algorithms, libraries, and techniques, whereas we see a gap for content and knowledge sharing on the actually building data science products at scale”, i.e., how to turn data sources and data science ideas into data science products in scalable and robust manner.
We would like to share with the data science communities and product companies our knowledge and experiences in building data science products. We want to learn building data science products at scale with other companies and data science practitioners.
Intended audience and takeaways
product manager: principles and practices for full product cycle from visioning to road mapping, spec’ing, and releasing
business leadership: deeper and concrete understanding of building data science products to enable product teams
data scientist, machine learning engineers: enlightenment of turning data into products that generate impacts for business
Speaker background
We are Data Science Product builders, hands-on in Product Managers, Data Scientists, Machine Learning Engineers, with shipped products including Mobility Intelligence DaaS, Web Intelligence, Telco Network Optimization, Behaviour Targeting and other ad targeting, Online Recruiting Platform, Privacy Management Systems, Personal Educational Assistant, etc.
Eureka AI builds data science products that enables partnerships between mobile operators and business partners to deliver relevance for consumers and revenue for partners. We bring ready-made solutions to telcos that can be efficiently deployed at speed to immediately accrue maximum value to their data. For consumer facing businesses we offer proven use cases to capture business opportunities and reduce risk.