Conference overview: "literally everything" from FinovateSpring 2017

Conference overview: “literally everything” from FinovateSpring 2017

Written by FinTech Futures
5th May 2017

Wished you were at FinovateSpring in San Jose but missed it? No worries – we’ve got it covered.

Kirsty Cooke, head of marketing at Mapa Research (Banking Technology‘s sister company), brings a comprehensive overview of the two-day conference – session by session.

This year, FinovateSpring focused on big data, machine learning and bringing investment to the masses. There was also quite a lot around digital mortgage experience.

FinovateSpring 2017 Day One – 26 April

We were welcomed by the usual pumping dance music and Erika Sanchez (head of events), who talked to the “whole new world” of financial services – the blossoming of biometrics, virtual reality, AI and blockchain. She told us to expect presentations from 59 of the “hottest companies in fintech” spanning the entire industry.

Session One (morning)

Ephesoft (who was also at FinovateEurope) was presented by Alex (VP) and Ike (CTO, but also playing the role of criminal) who talked about how Ephesoft Insight can extract meaning from unstructured data (that is, document content). Through their content analytics platform, you can “accelerate meaningful transactions and reduce fraudulent ones” or “use document content to catch bad guys”. Using (supervised) machine learning and big data (first buzzword klaxon!) they have already helped the intelligence industry and now plan to help the financial services industry, especially trade finance as it is the perfect target: it is complex, there are many parties involved, and lots of documentation containing that juicy unstructured data.

But, as they demonstrated, their platform can not only “read” these documents (like invoices), they can visualise unusual occurrences like over and under invoicing (deviation from a price index), where shipments are occurring to high risk destinations, or the activities of particular companies (with what they call a mind map and predictive analytics). Had drinks with these guys and they were excellent company!

John and Glenn from BancPass were up next, introducing the arguably slightly niche PToll App, “leveraging” (sorry, I hate that word but it’s unbelievably common at events like this) their API to “transform the North American toll market” and modernise toll system software. They claim access to their API will improve the customer experience and crucially reduce the marginal cost of toll transactions. Users can pay tolls from coast to coast (well, they’ll be able to by 2018) – an industry that apparently raises something like $15 billion… or five time this amount in 2030.

The sign-up processes uses (or can use) Google login credentials; the user can then choose a payment provider e.g. PayPal; can then take a picture of their car’s licence plate to complete registration. Usefully, they can add additional vehicles such as a rental car and associate charges for this vehicle with a business account – something my American colleague could see the benefit of. Users can see all the transactions they make in a feed and click on them to find out more about them, and there is also a messaging system.

Harry, Miron and Max from SuperMoney talked about the growth in online lending and the “typical borrower experience”: the customer googles “personal loans”, dozens of options appear, including loan aggregators. Most aggregators use a pingtree (?) model, offering a waterfall of options that really doesn’t take customer needs into account, they argue. Think about what Kayak is for airlines… they want to bring the “same great user experience” to financial services by guiding users to loan offers in their best interest.

There is a really nice form-filling experience – tied up with some Google APIs to make it quick and easy. They do a soft credit check, I believe, and the example brings up 22 offers from five lenders where the applicant is pre-qualified and often pre-approved for a loan. They can sort by the obvious stuff like APR, payment etc. and click to get more information e.g. all fees. In a nice touch, they can also see user reviews of the brand to give a ‘qualitative dimension’ to the process.

In the example, the user clicks through and everything has pulled through to the loan provider in question, Prosper. So they just need to sign up and prove ID. These guys are pushing the customer-centric Google approach of Relevance rather than Highest Bidder. And ‘Guess what! Personal loans is just our first vertical!’ so watch this space.

Jumio (who also presented at FinovateEurope), brought Robert (CRO) and Ervinna, the head of customer success. They argued that “the only practical approach to KYC for fintech is a hybrid of machine learning, tech and IT experts” saying that digital identity must connect to real-world identity (which is what they do.)

Jumio’s NetVerify performs 150,000 verifications a day, the presenters said. The very enthusiastic Ervinna tried to pretend to be Robert with a high-res print-out of his face and was of course foiled because they also “know your cheater”. Jumio works with “most large financial services organisations” already, apparently.

And the list goes on… Other companies and products featured in the review include:

Topayz
PayNearMe
Tyro Payments
Layer
Capitalise
Synapse Financial Technologies
Finicity
Eltropy
and more!

Click here to read the full article: FinovateSpring 2017 Day One

FinovateSpring 2017 Day Two – 27 April

A touch hungover, grasping a coffee and munching a seeded bagel like a boss, I came back after a brilliant Day One and once again enjoyed the banging dance music as host Erika introduced another day of presentations and the latest innovations in fintech…

Session One (morning)

We kick off with Capsilon Corporation who talk about “powering the digital mortgage factory”. COO Jim introduces the newest digital mortgage tech solution – enabling mortgage companies to move “from a labour-centric to technology-centric model” that will “accelerate overall process and transform the economics” of mortgages so you can “deliver quality mortgages at velocity”.

He starts the demo with a lender portal that has Capsilon tech embedded. On the dashboard, you register a loan and then Capsilon finds information in the loan documentation in order to create a checklist for things you’ll need i.e. required documents.

The technology finds the names of different documents from within larger files and sorts them automatically. Their solution reduces up to 80% of the labour associated with receiving, sorting, naming and filing these documents. The system also validates the digital assets received. From the underwriter perspective: they will receive the documents and use pre-set rules to find errors
Jim and Neil have a game-changer, they say. A totally new financial product category – Unison Home Ownership Investors invests in your home with you. The partnership model means a10% deposit from you and 10% from Unison. Rather than a demo of any technology, they show us the calculators on the website: letting you see what max value you can get, and which houses you can afford in the area you want. “Use our money for up to 30 years without interest,” they promise. Their solution is about “connecting homebuyers to patient capital” and “making real estate an investible asset class.”

Ana the CEO introduces Infocorp. For 20 years they’ve been “making people love their bank”. They do this by helping banks anticipate customer needs, make them feel special, and provide relevant offers. IC-Banking allows banks to show an offer in emails, mobile banking, desktop banking or/and social media, or from an Outlook tool. She talks about creating “omnichannel campaigns” that are “omni relevant”.

IC Campaign Manager lets you orchestrate these campaigns simply, from one place. You can define what happens when a user clicks the offer, and if he closes the offer, the platform stops showing the offer in all channels. So you won’t be “omni annoying” – lol. It uses Machine Learning to work out the next most relevant campaign and, if the customer converts, a new probability of conversion will have been calculated and a new campaign is shown.

Each advert can be personalised e.g. names and loan amounts and users behind the scenes get an analysis dashboard to see minute by minute campaign performance and what customers are seeing and doing.

Tim CEO and co-founder introduces BeSmartee by saying that mortgages suck and are complicated. We were mortgage veterans, he says, and saw what 2008 “did to communities”. He introduces their point of sale solution – the Smart Mortgage application process. It includes continuous data validation e.g. real phone number required on the form. Customers enter a zip code, property value, and deposit amount – the system then finds options based on this data – then some more borrower details are required, including SSN. Payroll info and bank data is imported, and they “merge big data with artificial intelligence behind the scenes” apparently. The submitted loan application gets a “patented 10 0 3 merge” treatment, whatever that is. The initial disclosures are sent and can be e-signed – without taking the customer off to a third party. There is then a loan dashboard for customer communications and so on, and a similar dashboard for lenders.

Fuco& Solution

Taiwan-based Fuco&Solution presented an “innovative mobile statement solution”. Apparently, 100 million statements every month are generated by them. As branches close, they say, the statement becomes an important comms strategy – and now they have invented “interactive, robo-advisory” statements. From these, customers can pay a credit card bill, drill down into transactions, split payments (pay in instalments) and see offers based on spending habits. It can be integrated with products I can buy with loyalty points, I can see and purchase funds, and if I decide to take e.g. a mortgage offer, I can do a quick comparison and application right there and then. These guys let banks “monetise credit card statements” and the whole thing is configured (although we didn’t see this) with “drag and drop building blocks”.

Jeff, CEO of Neener Analytics, told us that 56% US consumers are credit challenged. In order to expand the market, where risk is unknown or unreliable, he presented a solution that “can reduce risk and increase volume and revenue with a 20-minute application”. It is also a regulatory-compliant solution. This is “social media decisioning analytics that actually work”. Not based on who we are friends with; and not “can they pay back?” but “will they?”. He talks about the “biometrics of personality”.

A customer can apply for credit, hit login with Facebook, then continue with the application i.e. it takes just one click for customers. Meanwhile, in the background, an ‘automated psychologist’ is looking at 100s of data points and combining these into a risk profile. Users can see the business dashboard (dials) which help with identity verification, measure ‘Veracity’ and ‘Resilience’ and predict the likelihood of default, revolving credit or paying off early. ‘Stop using the old data. Find something that actually works.’

AutoKept was presented by a “rancher from Arizona” (a real character!) who can never remember what his expenses were. His intern (?) Travis said “give me four weeks” (to build a solution) and it’s taken two years. But the tool lets you take a picture of a receipt and record yourself saying what it was for. Then submit and it will either go to QuickBooks or a spreadsheet. It’s useful for transactions that take place at e.g. Walmart where it’s too generic a brand to remember what you actually bough (and folk just guess).

Talk Accounting: Take a picture. Talk about the transaction. Submit. That’s it!

The app takes data from the bank and accounting software, asking the user for the rest of the details. You verbally say what it was for and take a picture of the receipt. Travis showed us how the solution perfectly recorded him saying: “Let’s go turn the microphone on, I’m just going to say this was for a printer.” The rancher is “looking for companies that can send me the last four digits” (i.e. customer card data to tie it all together automatically).

And the list goes on… Other companies and products featured in the review include: