Entrepreneur First (EF) doesn't find startups: it builds them. The startup accelerator, which was founded in 2011 by Matt Clifford and Alice Bentinck, takes talented technologists and turns them into startup co-founders, all in just six months.
Today, EF's seventh group of 18 startups pitched to a packed hall of investors at King's Place in London. Just six months ago, they were strangers – but you would never have guessed it from the smoothly-delivered presentations.
The EF process takes place in two halves. First, there are three months of startup formation, in which co-founders meet and decide what to work on. At the end of that stage, the EF team decides whether or not to invest, taking an eight per cent equity stake in exchange for a £10,000 investment and access to a £70,000 convertible note.
Read more: Become an entrepreneur in four steps
Twenty-five startups made it through – a total of 50 people. (EF startups are always pairs.) In the second stage, the product-building stage, they were whittled down to 18. EF specialises in turning PhDs into entrepreneurs. The final teams had 181 research papers and 90 patents between them.
For me, it was especially interesting to see the teams, because I spent the last three months of 2016 following this cohort to see how the EF process worked. (My deep dive into EF will be published next month.) One lesson: the final result may appear easy, but the process is far from simple.
Here are the companies, beginning with my four picks for ones to watch.
Eyn
Eyn has built an AI-powered identity-checker, using facial recognition algorithms to turn smartphones into biometric ID verification devices. Co-founded by Mohamed Ben Arbia and experienced security researcher Hasan Sheikh Faridul, this team formed early and had a functioning prototype after only three months.
Scape
If augmented reality is going to work, it'll need a system like Scape. This startup makes it possible to overlay digital content on the physical world, by detecting the exact location of the user's device, even if it's outside.
Like Eyn, the pair behind Scape – which includes VR industry veteran Edward Miller - formed early and worked well together. They have already secured a pilot with one of the world's largest theme parks.
Optimal
The impressively-credentialed pair behind Optimal - co-founder David Hunter was head of Algorithmic Trading at Deutsche Bank – moved to the Netherlands to observe greenhouses in action, and believe their AI-powered algorithms can change the way these traditionally-powered operations run.
First step: reduce greenhouses' energy use. "We expect at least a 15 per cent reduction in energy," Hunter told the audience. But this, he said, was the "first step in a master plan" working towards "complete auto-pilot for greenhouses".
This pitch wasn't seen on the live-stream - a sign, in previous Demo Days, of a startup on its way to funding. Judging by the investor gossip, the same should be true of Optimal.
Marble
Marble co-founder Mathieu Johnson went all the way through EF6, only to have his co-founder drop out days before the final Demo Day. Now the former Airbus engineer is back, working with Charles Peurois – a Mercedes F1 designer he found outside the EF process - to build a fleet of high performance drones.
Sentient Machines
Not to be confused with Sentient.AI, the hugely-funded artificial intelligence firm founded in 2007 by Antoine Blondeau, this startup uses speech-to-text software to help call centres speed up their work. Its co-founders - Danica Damlijanovic and Lorenz Fischer - both have deep learning PhDs.
Oilfront
OilFront helps shipping companies buy fuel directly from suppliers, instead of going through brokers, a market it says is worth $200 billion per year. CEO Danny Soos described it as "bunker procurement software for the shipping industry" (bunkering is the shipping jargon for the supply of fuel).
Scalia
"71 per cent of e-commerce product listings have incomplete, inaccurate or wrong information," co-founder Lancelot Salavert told the audience. The solution? A unified product platform that uses machine learning to map products, potentially putting all e-commerce firms on the same data standard.
Observe Technologies
Calling itself "the world's first artificial intelligence platform for fish farming," Observe Technologies looks to replace the current "manually-intense" process with automated systems. Applying computer vision and machine learning to simplify the feeding process, its product gives fish farmers intelligent recommendations about how fast and how much to feed their stock.
One sign of possible progress: a deal with one of the UK's biggest fish producers that gives Observe a database of five million fish images on which to train its algorithms.
Transformative
Transformative predicts heart attacks before they happen. Its algorithm, cardioAI, is able to predict the onset of sudden cardiac arrest (although details, at least in the pitch, remained vague). The team, including a former CERN researcher, has an agreement with Northern Estonia Medical Centre to carry out clinical trials.
Proportunity
Proportunity uses machine learning to accurately forecast house prices and up-and-coming areas. (Yes, it's an AI estate agent - just what we always wanted.) A big claim, but the startup has plenty of data: it's working with three of the top five estate agencies, including Savills, which has invested in the company.
Multiply
An AI-powered financial advisor, Multiply claims to be able to tell ordinary consumers everything they should do with their money, from mortgages to investments to savings. "We've developed a revolutionary technique to generate advice," says co-founder Vivek Madlani, asking the audience of investors for £1 million in seed funding.
Lendr
A rare EF consumer startup, Lendr is a reverse auction platform for mortgages - a system that competes against existing brokers by using a mass of data parsed by, you guessed it, artificial intelligence. The founders came together early and are already working with five major lenders and the FCA before a summer launch.
Plotist
Plotist is a software aid for narrative, helping would-be storytellers weave together their plots, characters and themes. Co-founder Jay Cano developed the idea himself, before coming to EF to find a partner - all-round developer Simon Altschuler.
Creative AI
Need more Facebook ads in your life? Creative AI's machine-learning powered marketing is designed to stand out on social media. Co-founders Lukas Dirzys and Yazan Al Alami combine private data sources (Facebook page, product inventory) with publicly available information (news, trending topics) to create their secret sauce.
ThinkSono
Deep vein thrombosis (DVT) - scourge of airlines, and killer of 800,000 people a year, more than breast cancer, prostate cancer, AIDS and car accidents combined. That's according to ThinkSono, who have built software that allows any healthcare professional to diagnose DVT at the point of care using a portable scanner. Co-founder Fouad Al-Noor specialised in medical imaging in his nanotechnology master's at the University of Southampton.
Predina
Predina Technologies mitigates risk in autonomous vehicles, using a whole range of contextual data to predict how likely they are to have an accident. Think of it as insurance for driverless cars.
Predina CEO Bola Adegbulu has impressive domain knowledge: he used to manage dealer relationships for Jaguar LandRover. His co-founder Meha Nelson, a machine learning specialist, provides the tech heft, but she almost didn't make it: Nelson spent weeks during the first stage working on a speech-to-text product, only to realise it wasn't going to scale.
Selerio
This brother and sister team joined EF expecting to work separately - then found themselves teaming up to found Selerio, an advertising startup that claims to be able to put "real-time, targeted, product placements" directly into videos. Already working with ITV, Flora and Ghislain Tasse, Selerio has an impressive background, plus that unbeatable sibling chemistry.
Phoelex
Each Demo Day, EF springs one surprise: a solo entrepreneur who didn't work with the group, but makes an appearance all the same. Today's was Amir Nejadmalayeri, an award-winning scientist with over 40 publications, whose startup Phoelex promises to "revolutionise" data centres with transceivers that are "5x cheaper, 10x smaller, and 3x more energy efficient." Huge, if true.
This article was originally published by WIRED UK
