KDnuggets Home » News » 2014 » May » Opinions, Interviews » Interview: Dale Russell, CTO, Talksum on Building Talksum Router and Real-time Anaytics ( 14:n13 )

Interview: Dale Russell, CTO, Talksum on Building Talksum Router and Real-time Anaytics


We discuss challenges in building Talksum data stream solution, current trends in real-time analytics, advice for Data Science aspirants and more.



Dale RussellDale Russell, CTO, Talksum, Inc., has over 20 years of front-line applied engineering and operations management expertise. Having managed large-scale infrastructures and service delivery for major names such AOL and Salesforce.com, he has an in-depth knowledge of the enterprise technology landscape. Dale has also focused on large-scale data management and filtering technologies for large advertising networks and security companies, and his vision for the next era of computer science innovation is at the heart of the Talksum Data Stream Router solution.

Part 1 of interview: Dale Russell on Winning the IE Big Data Startup Award

Here is part 2 of my interview with him:

Anmol Rajpurohit: Q4. What were the biggest challenges that you had to overcome while building the Talksum Data Stream solution? Were there any interesting observations during the solution development that significantly impacted your solution design/architecture?

Talksum routerDale Russell:
The hardest hurdle has been changing people perceptions of “Big Data” and showing them it is possible to overcome the perceived challenges from the words “The Big Data Problem”.
We have a recurring question from most new customers and even new employees, “What language do we have to learn to integrate your API into our existing system?” The second question is, “Does this mean we have to send our data to a cloud?” The industry has seen solutions wrapped in an API for so long that now the public seems shocked by a “Turn-key” solution. Once they realize we’re not kidding, we usually see a smile on their face.

AR: Q5. What motivated you to switch your career from applied engineering and operations management to large-scale data management? How did you got involved with Talksum?

DR: I do not see this as a career change; my career has been one of solving problems. The point of applied engineering is about implementing new technologies to solve a particular problem. Operations management is about engineering sanity: knowing which technologies solve problems and which technologies will lead you down the proverbial rabbit hole. Talksum is the greatest extension of the two careers. Talksum provides a reliable—and almost boringly dependable—solution for the “Big Data" problem. I have been tremendously fortunate to work in operations at some of the largest data centers and service delivery organizations, and the Talksum Data Stream Router is the byproduct of need.

AR: Q6. Which of the current trends in Real-time Analytics are of the most interest to you? How do you see things change in next few years?

Big DataDR: This may be the hardest question. While do I see a lot of great trends in Real-Time Analytics, I do not see a cohesive strategy. A lot of great work is going into simple languages for describing run time algorithms. The storage vendors bringing in some of the more traditional NoSQL data stores into their fabric is pretty cool. I think over the next few years we see a larger shift to focusing on how to gain benefit from our data rather than how we continue to scale overly complex application tiers. Talksum is focused on viewing your data flows as an integral part of your infrastructure rather than an application burden.

AR: Q7. Based on your experience, what advice would you offer to people aspiring a long-term career in Data Science?

AdviceDR: Lift your head up occasionally, observe one's surroundings.
Too often we hyper-focus on the wrong things as engineers or scientists, and we tend to fall in love with our code, our process, or our tools. Instead, we should focus on the problem, not the trendy library we hear about that others are using.
Gather specs, know these specs will be wrong or incomplete, and focus on solutions that are forward-looking so you can focus on proving meaningful insight to the data, instead of on simply writing ETL scripts.

AR: Q8. On a personal note, are there any good books that you have been reading lately and would like to recommend? Chaos book

DR: If you have not read James Gleik's "Chaos, Making a New Science" you should. Another favorite is "The Never-Ending Days of Being Dead: Dispatches from the Front Line of Science" by Marcus Chown. My family says I have a pretty boring book selection: all Science, Political Science, or stacks and stacks of comic books.

Related: