[TALK]@Telematika

Talks & Presentations Collection

Securing Apache Kafka
December 10, 2019

Securing Apache Kafka

With Apache Kakfa 0.9, the community has introduced a number of features to make data streams secure. Jun Rao explains the motivation for making these changes, discusses the design of Kafka security, and demonstrates how to secure a Kafka cluster. Jun also covers common pitfalls in securing Kafka and talks about ongoing security work.

Semantic natural language understanding with Spark Streaming, UIMA, and machine-learned ontologies

Semantic natural language understanding with Spark Streaming, UIMA, and machine-learned ontologies

December 10, 2019

David Talby and Claudiu Branzan lead a live demo of an end-to-end system that makes nontrivial clinical inferences from free-text patient records. Infrastructure components include Kafka, Spark Streaming, Spark, Titan, and Elasticsearch; data science components include custom UIMA annotators, curated taxonomies, machine-learned dynamic ontologies, and real-time inferencing.
Statistics and the art of deception

Statistics and the art of deception

December 10, 2019

Data scientists use statistics to reach meaningful conclusions about data. Unfortunately, statistical tools are often misapplied, resulting in errors that cost both time and money. Deborah Berebichez presents examples of egregious misuses of statistics in business, technology, science, and the media and outlines the simple steps that can reduce the chance of being fooled by statistics.
Stream analytics in the enterprise: A look at Intels internal IoT implementation

Stream analytics in the enterprise: A look at Intels internal IoT implementation

December 10, 2019

Moty Fania shares Intels IT experience implementing an on-premises IoT platform for internal use cases. The platform was designed as a multitenant platform with built-in analytical capabilities and based on open source big data technologies and containers. Moty highlights the lessons learned from this journey with a thorough review of the platforms architecture.
The future of column-oriented data processing with Arrow and Parquet

The future of column-oriented data processing with Arrow and Parquet

December 9, 2019

In pursuit of speed, big data is evolving toward columnar execution. The solid foundation laid by Arrow and Parquet for a shared columnar representation across the ecosystem promises a great future. Julien Le Dem and Jacques Nadeau discuss the future of columnar and the hardware trends it takes advantage of, like RDMA, SSDs, and nonvolatile memory.
The Netflix data platform: Now and in the future

The Netflix data platform: Now and in the future

December 9, 2019

The Netflix data platform is constantly evolving, but fundamentally it's an all-cloud platform at a massive scale (40+ PB and over 700 billion new events per day) focused on empowering developers. Kurt Brown dives into the current technology landscape at Netflix and offers some thoughts on what the future holds.
The new dynamics of big data

The new dynamics of big data

December 9, 2019

Since its inception, big data solutions have best been known for their ability to master the complexity of the volume, variety, and velocity of data. But as we enter the era of data democratization, theres a new set of concerns to consider. Mike Olson discusses the new dynamics of big data and how a renewed approach focused on where, who, and why can lead to cutting-edge solutions.
The personalization spectrum

The personalization spectrum

December 9, 2019

How are users meant to interpret the influence of big data and personalization in their targeted experiences? What signals do we have to show us how our data is used, how it improves or constrains our experience? Sara Watson explains that in order to develop normative opinions to shape policy and practice, users need means to guide their experiencethe personalization spectrum.
The tech behind the biggest journalism leak in history

The tech behind the biggest journalism leak in history

December 9, 2019

The Panama Papers investigation revealed the offshore holdings and connections of dozens of politicians and prominent public figures around the world and led to high-profile resignations, police raids, and official investigations. Almost 500 journalists had to sift through 2.6 terabytes of datathe biggest leak in the history of journalism. Mar Cabra explains how technology made it all possible.
Transforming healthcare through precision data science: Myths and facts

Transforming healthcare through precision data science: Myths and facts

December 9, 2019

Healthcare, a $3 trillion industry, is ripe for disruption through data science. However, there are many challenges in the journey to make healthcare a truly transparent, consumer-centric, data-driven industry. Sriram Vishwanath shares some myths and facts about data science's impact on healthcare.
Triggers in Apache Beam (incubating)

Triggers in Apache Beam (incubating)

December 9, 2019

Triggers specify when a stage of computation should emit output. With a small language of primitive conditions, triggers provide the flexibility to tailor a streaming pipeline to a variety of use cases and data sources. Kenneth Knowles delves into the details of language- and runner-independent semantics for triggers in Apache Beam and explores real-world implementations in Google Cloud Dataflow.
Tuning Spark machine-learning workloads

Tuning Spark machine-learning workloads

December 8, 2019

Spark's efficiency and speed can help reduce the TCO of existing clusters. This is because Spark's performance advantages allow it to complete processing in drastically shorter batch windows with higher performance per dollar. Raj Krishnamurthy offers a detailed walk-through of an alternating least squares-based matrix factorization workload able to improve runtimes by a factor of 2.22.
Turning petabytes of data into millions in cost savings for the worlds biggest retailers

Turning petabytes of data into millions in cost savings for the worlds biggest retailers

December 8, 2019

Jonathon Whitton details how PRGX is using Talend and Cloudera to load two million annual client flat files into a Hadoop cluster and perform recovery audit services in order to help clients detect, find, and fix leakage in their procurement and payment processes.
Unlocking unstructured text data with summarization

Unlocking unstructured text data with summarization

December 8, 2019

Our ability to extract meaning from unstructured text data has not kept pace with our ability to produce and store it, but recent breakthroughs in recurrent neural networks are allowing us to make exciting progress in computer understanding of language. Building on these new ideas, Michael Williams explores three ways to summarize text and presents prototype products for each approach.
US venture: Risk, values, founder outcomes

US venture: Risk, values, founder outcomes

December 8, 2019

Susan Woodward discusses venture outcomeswhat fraction make lots of money, which just barely return capital, and which fraction fail completely. Susan uses updated figures on the fraction of entrepreneurs who succeed, including some interesting details on female founders of venture companies.
Using graph databases to operationalize insights from big data

Using graph databases to operationalize insights from big data

December 8, 2019

Tim Williamson and Emil Eifrem explain how organizations can use graph databases to operationalize insights from big data, drawing on the real-life example of Monsantos use of graph databases to conduct real-time graph analysis of the companys data to transform the business in ways that were previously impossible.
Virtual Cove

Virtual Cove

December 8, 2019

.
VoltDB and the Jepsen test: What we learned about data accuracy and consistency

VoltDB and the Jepsen test: What we learned about data accuracy and consistency

December 8, 2019

VoltDB promises full ACID with strong serializability in a fault-tolerant, distributed SQL platform, as well as higher throughput than other systems that promise much less. But why should users believe this? John Hugg discusses VoltDB's internal testing and support processes, its work with Kyle Kingsbury on the VoltDB Jepsen testing project, and where VoltDB will continue to improve.
Watermarks: Time and progress in Apache Beam (incubating) and beyond

Watermarks: Time and progress in Apache Beam (incubating) and beyond

December 8, 2019

Watermarks are a system for measuring progress and completeness in out-of-order streaming systems and are utilized to emit correct results in a timely manner. Given the trend toward out-of-order processing in existing streaming systems, watermarks are an increasingly important tool when designing streaming pipelines. Slava Chernyak explains watermarks and explores real-world applications.
What Crimean War gunboats teach us about the need for schema registries

What Crimean War gunboats teach us about the need for schema registries

December 8, 2019

In 1853, Britains workshops built 90 new gunboats for the Royal Navy in just 90 daysan astonishing feat of engineering made possible by industrial standardization. Snowplow's Alexander Dean argues that data-sophisticated corporations need a new standardization of their own, in the form of schema registries like Confluent Schema Registry or Snowplows own Iglu.
A peek at x.ais data science architecture

A peek at x.ais data science architecture

December 7, 2019

In any human-machine interaction, you need a dialogue model: the machine must understand and be able to respond appropriately. Angela Zhou discusses x.ai's AI personal assistant, Amy Ingram, who schedules meetings for you, focusing on the way x.ai has approached both understanding and responding.
AI is not a matter of strength but of intelligence

AI is not a matter of strength but of intelligence

December 7, 2019

Francisco Webber offers a critical overview of current approaches to artificial intelligence using "brute force" (aka big data machine learning) as well as a practical demonstration of semantic folding, an alternative approach based on computational principles found in the human neocortex. Semantic folding is not just a research prototypeit's a production-grade enterprise technology.
Artificial intelligence: Making a human connection

Artificial intelligence: Making a human connection

December 7, 2019

Genevieve Bell explores the meaning of intelligence within the context of machines and its cultural impact on humans and their relationships. Genevieve interrogates AI not just as a technical agenda but as a cultural category in order to understand the ways in which the story of AI is connected to the history of human culture.
Benefits of scaling machine learning

Benefits of scaling machine learning

December 7, 2019

Machine learning is evolving to utilize new hardware, such as GPUs and large commodity clusters. Reza Zadeh presents two projects that have benefitted greatly through scaling: obtaining leading results on the Princeton ModelNet object recognition task and matrix computations and optimization in Apache Spark.
Building and applying emotion recognition

Building and applying emotion recognition

December 7, 2019

Anna Roth and Cristian Canton walk you through building a system to recognize emotions by inferring them from facial expressions. Cristian and Anna explain how they trained their emotion recognition CNN from noisy data and how to approach labeling subjective data like emotion with crowdsourcing before showing a demo of this work in action, as it is exposed in Microsofts Emotion API.
Unlock the power of AI: A fundamentally different approach to building intelligent systems

Unlock the power of AI: A fundamentally different approach to building intelligent systems

December 7, 2019

Mark Hammond explains how Bonsais platform enables every developer to add intelligence to their software or hardware, regardless of AI expertise. Bonsais suite of toolsa new programming language, AI engine, and cloud serviceabstracts away the lowest-level details of programming AI, allowing developers to focus on concepts they want a system to learn and how those concepts can be taught.
What ties to what? Visualizing large-scale customer text data with bipartite graphs

What ties to what? Visualizing large-scale customer text data with bipartite graphs

December 7, 2019

Which suppliers are most likely to have delivery or quality issues? Does service, product placement, or price make the biggest difference in customer sentiment? Text data from sources like email and social media can give answers. Mark Turner explains how to see the associations between any two variables in text data by combining text analytics and the bipartite graph visualization technique.
When one data center is not enough: Building large-scale stream infrastructures across multiple data centers with Apache Kafka

When one data center is not enough: Building large-scale stream infrastructures across multiple data centers with Apache Kafka

December 7, 2019

You may have successfully made the transition from single machines and one-off solutions to large, distributed stream infrastructures in your data center. But what if one data center is not enough? Ewen Cheslack-Postava explores resilient multi-data-center architecture with Apache Kafka, sharing best practices for data replication and mirroring as well as disaster scenarios and failure handling.