Future Focus: Innovation BreakfastTue 16 Apr 2019
Last week we held the first of our Future Focus events – showcasing innovations from across the market.
I say ‘our first’, because we’ve had a lot of requests to do another, and great feedback from the first – thanks go to all the speakers! Particular thanks to Derk Louwerse, CEO of Shell Fitcar, who opened with a well-received keynote.
If you couldn’t make it to the event, read my summary below and come back soon as we will upload videos of most of the presentations here on www.lloydslab.com/news.
Using HoloLens to better visualise risks
The event was all about innovations which are actually happening in the market, so it was great to have Bobbie Mansfield from AXA XL bring a HoloLens headset and show us how they have been using it to bring to life something which can be hard to understand or visualise.
They are using the tech to simulate real-life scenarios, for example with the positioning of sprinklers in warehouses. Are they in the right place? What’s changed since last year?
The units are heavy and have poor battery life, but as competition grows, AXA XL plan to be early adopters in this space.
Lee Timms from BeazleyLabs shared their 8-week experience of automating part of the claims process – accepting the small hit they would take from leakage.
It was interesting to hear about the challenges they faced with data quality, completeness, and volumes. By pivoting early on, collaborating, and building a ‘bag of words’, they were able to develop an ML solution which added value.
Adam Mitchell from Chaucer shared their experience with Praedicat and Arium – leveraging activities which kicked-off here in the Lloyd’s Lab.
Can we trace salmonella outbreaks back to the specific animal? Which theoretical risks are going to be something to really worry about in the future?
By utilising external start-ups, Chaucer gained better insight into their current portfolio, and can make more informed risk decisions.
Steven Wilkins from Hiscox talked to us about how they have used analytics to improve decision-making, upskill sales teams, and drive revenue increases.
Their 4-stage process and use of CRAN helped them develop a solution which can predict what products a customer is likely to need, and the amount of cover.
It would have been great to have seen the real thing, but Chris Lovick at MS Amlin gave a great overview of their Alexa dot demo which they built in just a couple of days – showing how there is a lot of tech out there which you can quickly leverage without reinventing.
A pitch to the board, with no PowerPoint, and a voice command to linked to the DVLA – “try things out, without fear”!
Jamie Garratt from Talbot talked about a platform they built to improve distribution for US political violence risks. Quickly going from nothing to 7,500 quotes and millions in profitable business, it was a great example of how to move quickly.
Jamie finished with three lessons they learned:
- Experiment and be ready to make mistakes
- You don’t need to start from scratch (as Chris demonstrated above!)
- Talk to your customers
Lloyd’s data lab (different to the Lloyd’s Lab!)
AI, IoT, and big data are three areas that Craig Civil from the Lloyd’s Data Lab are exploring.
Craig gave an intro to an NLP solution they’ve built on top of Crystal with some impressive results – AI-driven answers to queries in 45 seconds, rather than over 6 minutes for a human!
Lookout for more from them regarding NLP on unstructured, OCR‘d documents.
That’s all for now.
Here are some quick explanations for the tech terms used above:
|Big data||Lots of data – often defined by Velocity (how quickly new data comes in), Volume (how much data you have), and Variety (how many different types)|
|CRAN||The Comprehensive R Archive Network. R is an open-source software for statistical computing and graphics|
|HoloLens||Microsoft’s mixed reality headset.|
|IoT||Internet of Things. e.g. your phone, or smart TV|
|ML||Machine learning – a type of artificial intelligence|
|NLP||Natural Language Processing – using AI to understand the meaning of written words|
|OCR||Optical Character Recognition – e.g. getting text from scanned documents|