I attended the Hitachi Vantara DataOps NEXT Virtual Conference for the third year in a row, albeit virtually for the first time. I miss their big conferences, which are meticulously planned and executed. I listened in for one full day, seven or eight presentations.
Some background: Hitachi Vantara is a hardware, software, and services organization, now that Hitachi Vantara merged with Hitachi Consulting to accelerate the expansion of Hitachi's Social Innovation Business and Digital Growth. January 2020 was the inauspicious kick-off date for the integrated companies.
Hitachi Consulting was formed in 2000 when the Japanese multinational corporation Hitachi Ltd. acquired Grant Thornton LLP's Chicago-based information technology and strategy consulting practice. Other significant acquisitions include 23 partners and 370 consultants of Arthur Andersen Business Consulting (AABC), which joined the company in June 2002 after the dissolution of Arthur Andersen. I don't have precise numbers for Hitachi Consulting, but it was 6500 in 2015. Hitachi is an $85 billion company, with roughly 300,000 employees worldwide.
Hitachi Vantara is a sizeable organization. There were four tracks at dataOPS NEXT this year:
- Optimise Data Fabric
- Build and Manage Data Pipelines
- Improve Data Governance and Agility
- Expand Analytics and Machine Learning
Notice that three of the four tracks contain the word "data." The concept of DataOps was everywhere. This is a departure from last year, where the hardware component of Hitachi Vantara was prominent. The other topic last ear was Sustainable Society through its Social Innovation Business. Toshiaki Higashihara, President and CEO of Hitachi, Ltd., gave a keynote on the subject.
The other keynote presentation was by Zeynep Tufekci, a Turkish writer, academic, and techno-sociologist, known primarily for her research on the social implications of emerging technologies such as artificial intelligence and big data - in the context of politics and corporate responsibility. But this year, social responsibility was not the main topic.
Lumada takes center stage
Lumada is software. Its focus is on analytics, governance, and operational agility. Anything that Hitachi Vantara does in software is within the Lumada umbrella. For the third year in a row, I was impressed with Lumada's capability in IoT, especially IIoT, which makes sense - Hitachi is a leading manufacturer of industrial equipment, and operates over 100 of its own factories.
Acquisitions are rounding out the Lumada portfolio, and they set the task of integrating the products and moving them forward. Pentaho, acquired in 2015, was primarily an ETL/ELT product based on an open-source model. When I saw it last year, I was disappointed, mainly with the use of the word "catalog," which they did not have.
However, another critical acquisition was Waterline Data, which brought AI-driven data discovery, first-class catalog features, and a semantic model. Some of these features were presented as available now, but on the roadmap for much more development. Fred Schults, Director of Product Management, AI/ML, said that Waterline is the "glue" that pulls everything together, filling a gap in their Edge-Core-MultiCloud proposition.
This slide was particularly useful to convey the direction they're taking:
On the left are dashboarding and visualization, which are a little underwhelming. For something as visible to executives and a broad cross-section of an organization, they need to up their dashboarding game. The market is full of vendors with better capabilities and semantic models driving them. On the right side, on the other hand, Lumada has AI/ML development and deployment and model health management, with "transparent AI," also something that is road-mapped. They claim that a single semantic model pulls it all together, and from what I've seen from them before, I don't doubt it.
Many vendors claim to have semantic models, but I did learn that Lumada's also includes relationships, which implies a knowledge graph. That is critically important to navigate data and expose hidden meanings.
Alex Gorelick, the founder of Waterline (who is no longer with Hitachi), has been working on this problem since his first company, Acta, over twenty years ago. I know Alex, and I'm sure this is real. Lumada presenters, on several occasions, said that data science and ML processes typically take thirty or more applications and tools, but Lumada can deliver that in just three. I can't verify that, but in general, Hitachi Vantara has more of an engineering feel than a hungry start-up full of hype.
It's unfortunate that they laid out big plans last year and merged Hitachi Vantara and Hitachi Consulting just in time for the world to draw inward in a pandemic. No one I spoke to complained about this or offered any data about business activity, but it has to impact their momentum.
In the meantime, Hitachi Vantara has assembled a nearly complete set of offerings for DataOps, IIoT, AI/ML life cycle, and a platform for application development. Some of these things are on the "roadmap." However, it was hard to distinguish which were and which weren't. There would have been opportunities to get clarity in a conventional conference, but everyone is working under the limitations of the situation.