Hi Dave,
The Samoa piece is a bit tricky, the goal essentially is to take their storm 
components and enhance them to work within the heron storm subcomponent and 
eventually with heron streamlet architecture.  We chose Samoa because they have 
already built several machine learning topologies within storm.

Cheers

Sent from my iPhone

> On Jun 8, 2018, at 5:27 PM, Dave Fisher <[email protected]> wrote:
> 
> 
> 
> Sent from my iPhone
> 
>> On Jun 8, 2018, at 5:08 PM, Ning Wang <[email protected]> wrote:
>> 
>> Brief notes for today's meeting:
>> 
>> - Review DD:
>> https://docs.google.com/document/d/1LrO7XRcMxJoMM83wjRd-Ov74VAaomA_mXOAhCStgGng/edit
> 
> The document says copying Samoa. Heron should be working with the Samoa team 
> and being careful not to fork.
> 
> 
>> - We want to understand better about the bigger picture of ML in stream
>> processing systems.
>> -- talk to ML users
>> -- doc of related systems to read:
>>   ---
>> https://mapr.com/blog/monitoring-real-time-uber-data-using-spark-machine-learning-streaming-and-kafka-api-part-2/
>>   ---
>> https://databricks.com/blog/2018/06/05/introducing-mlflow-an-open-source-machine-learning-platform.html
>>   --- https://eng.uber.com/michelangelo/
> 
> In addition to talking to the vendors who are powered by Apache Spark, 
> directly talk to Apache Spark and Apache Kafka.
> 
> My 2cents.
> 
> Regards,
> Dave

Reply via email to