3 minute read
BDLdn is the largest Big Data Exhibition in the UK and an essential event for all data scientists to visit.
It was laid out as a large exhibition hall with 130 technology vendors and consultancies on stands, with several Seminar or presentation spaces scattered throughout the hall.
The Seminars were the big attraction with data gurus, or at least CTO level presenters from the major sponsors.
The Seminars were recorded, and you can subscribe and watch them on the organisers YouTube channel: https://www.youtube.com/c/Bigdataldn.
Presentations were either case studies of implementations, like McLaren from F1, commentary on industry trends, like the impact of real time, streaming data, or straight product pitches. Take your pick.
What Big Data London feels like
The seminar audience members were wearing headphones to cancel the background noise of the crowd in the exhibition all around them.
Not that the background noise was much of a problem, outside the seminars, which were well attended, the exhibition stands were often not crowded. I only once had to wait for more than 5 minutes to talk to a sales guy.
Mostly they jumped on me as soon as I stepped into the stand! Attendance was down from 2019. People are still concerned about COVID, I guess.
With 130 vendors present, every step of the journey from raw data to business information was represented.
Vendors at Big data London 2021
There were ETL vendors like Talend; database and data platform companies like MongoDB and Snowflake (though they were represented by partners, not with their staff, strange); there were system, data management, data governance tools; and front-end BI and Data science tools like Tableau and Tiger Graph.
What surprised me was there must have been 20 SME consulting firms. Finding expertise is still a challenge, possibly the biggest constraint to success with data science initiatives. The consultants were some of the more crowded booths.
If I could see a few themes, they would be:
- Responding to the requirement for streaming, or real-time, or IOT data. Several presentations were about using streaming data to provide insights into the health of operational processes.
Land Rover has implemented machine learning models across their supply chain with 4.5 billion components.
The attitude was 2020 was about reducing time to insight from 3 weeks to 45 minutes; 2021 is about reducing the response time from 45 minutes to milliseconds. To do that everything (data, the machine learning models,) have to be in memory.
Interestingly, Datavid is working on our first IOT data integration project. - The lack of skilled and experienced resources.
When talking shop all the consultancies and vendors admitted they are struggling to find both Experts and project level developers. The market is red hot. - The perennial search for an architecture that will allow “build once, reuse everywhere”.
A current frontier of the “build once, reuse everywhere” challenge is that Business Analysts and other BI users are typically using pre-defined data warehouses, whereas most data scientists are using their own “sandpits”.
The search for trusted data
The search for trusted data for all enterprise uses and users goes on still, the golden record or MDM is still elusive in the face of product innovation and the urgency of end-user requirements for insights.
The history of IT is that in each generation of product innovation the best of breed tools have eventually lost out to the platform vendors as customers simplify their system landscapes (and reduce their support costs).
All I will say is that I thought the most impressive demos I saw were from the specialist vendors for data scientists, Tiger Graph, DataIKU, and Data Robot, all deserving a mention here; but then I have always loved products with great UIs and flashy demos.
If the proliferation of tools is causing some confusion the other major issue I take away is that one of the major constraints to wider adoption is expertise.
A great experience all around
Business users have invested in intelligent data and remain willing to do so if data analysis begins to deliver some of the promised business value. Datavid as a SME specialist data services company is in a sweet spot in the market!
I can’t finish without thanking Andy Steed, the Content Manager at the organisers, 3rd Street Group.
Not only was the event well run but he was personally helpful to me.
When I shared I was suffering from “too much new information” syndrome, which frequently overcomes me at technical exhibitions he introduced me to Mike Ferguson of Intelligent Business Strategies who spent ten minutes giving me some context and wisdom. His presentation is also worth a viewing.
I did not even buy him a coffee!
A big thank you to Andy and Mike.
