EFTA00803391.pdf

HIVEMIND EFTA00803391  Software to build, clean and enrich data sets. We specialise in generating structured data from an unstructured world. Datasets as diverse as the content of documents or videos, group opinions, or the likelihood of future events. We believe humans and computers complement HIVEMIND each other powerfully, and that together they can tackle problems neither can complete alone. EFTA00803392  The problem Data-driven business decisions require structured data. 80% of data is unstructured', much more potential data is unrecorded. Transformative insights remain inaccessible to businesses within this raw material. Refining it into valuable data assets is hard to do flexibly, accurately and at scale. 1 Gartner: EFTA00803393  The solution Hivemind co-ordinates cognitive and automated processes to allow you to unlock the insights hidden in unstructured or unrecorded data. Break down big problems into bite-sized questions of parsing and judgement Distribute them either to humans or machines as appropriate Intelligently aggregate the responses to build data sets and ensure data quality Chain tasks together into sophisticated workflows to tackle hard data problems EFTA00803394  The impact Create bespoke datasets pertinent to your problems Transform your internal documents intq valuable assets I Clean and enrich your existing structured data Systematise manual data workflows and reduce senior staff time on basic data tasks Predict future outcomes by aggregating the knowledge of global experts or your own staff EFTA00803395  The product 0 A cloud-based platform with two interfaces: one for task creation, one for task completion. Hivemind Studio Hivemind Workbench Create, monitor, download Collect EFTA00803396  Target markets Financial research groups and the Data management operations Cross sector applications: legal, data vendors feeding that research across all financial firms medical, retail, tech - and many others 250 hedge funds > .$51)AUM Annual spend on financial By the end of 2O19, 9O% of large -200 asset managers >$5Ob AUM market data and news > $27b organisations will have a Chief Data -250 "alternative" & market data vendors Officer ( ) EFTA00803397  Business model Low barrier to entry. Cost scales with time, users, features and support. A Enterprise features Dedicated support Custom deployment Unlimited use 240+ - 1000 hr increase Annual Pro features Dedicated support £5000 p/a 120 License Fee (£000s) 24 Core Pro Enterprise O - Lite 250 5,000 25,000 Unlimited Usage (hours) EFTA00803398  HIVEMIND so far (0) At Winton... Since spin out In the press... 4 4 YEARS IN OPERATION WEEKS IN OPERATION The Economist FT 30+ BESPOKE DATASETS 14 QUALIFIED PROSPECT + Live... 8.5 million+ Credit Suisse Prime Services Conference 2017 COMPLETED TASKS Morgan Stanley European Quantamental Conference 2018 SBAI s Artificial Intelligence Roundtable 2018, New York 175+ 6 Alpha FMC & Illuminate Financial Fintech Showcase, 2018 TRIALS UNDERWAY The Technical Analyst's Alternative Data 2018 Event (forthcoming) \ MAN-YEARS OF WORK SBAI s Artificial Intelligence Roundtable 2018, London (forthcoming) Manager Magazin Awards Event 2018, Frankfurt (forthcoming) „..) EFTA00803399  Operational Challenges Challenges Distribution channels • Currently no distribution partners Competition Burden of consultancy Human as parser • Learning curve of using Hivemind • Figure Eight • Amazon Mechanical Turk Finance and the cloud • How do we convince a naturally Robotic process automation conservative industry to embrace a • Automation Anywhere cloud-based solution? • WorkFusion Workforce provision Human as expert • Develop partnerships with work- • Lumenogic outsourcers (e.g. Cloud Factory) and • Consensus Point improve integration with crowd source environments (e.g. Mechanical Turk) EFTA00803400 What makes HIVEMIND different? CORE CONCEPT SOPHISTICATION • Built for complicated workflows and • Focus on using computers to help dense, heterogeneous data sources humans solve hard problems more easily, • Sophisticated aggregation of human rather than on trying to replicate or judgement and expertise replace human intelligence. • Validated through four years of use by demanding data organisation FLEXIBILITY FUNDAMENTAL DATA VALUES • Not tied to a specific workforce • A practical solution for both big projects • Deals with wide-range of use cases and daily workload across a data scientist's workload • No reliance on buzzwords • Simple integration with a client's internal • Data quality is more important than workflow data size EFTA00803401  Team CEO DATA SCIENCE ENGINEERING Dan Mitchell Christian Gilson Alex Dawes • Director, Head of Research Data @ Winton • VP, Data Scientist @ Winton • VP, Technology @ Winton • Prev. Oxford Uni • Prev. Oxford Uni Mark Roulston Alex Taroghion CHIEF REVENUE OFFICER • MD, Research @ Winton • VP, Engineering @ Winton • Prev. Met. Office, Penn State Uni, • Prev. RBS, Barclays Henrik Grunditz Caltech, Cambridge Uni • SVP, Bus. Dev. @ Winton Riaz Karim • Prev. MSCI, Accenture, Imperial College • VP, Engineering @ Winton • Prev. JP Morgan, Goldman Sachs The team has worked together at Winton and on Hivemind over the last 4 years EFTA00803402  Projected financials (£000s) (O) EOY 20181 EOY 2019 EOY 2020 EOY 2021 EOY 2022 Enterprise Clients (- 240k) 1 2 3 5 12 Professional Clients (- 120k) 2 4 10 25 65 Core Clients (- 24k) 3 6 16 38 100 Revenue 240 906 1830 5112 13080 Revenue Growth (%) 0 139 102 179 175 Total Costs -562 -2416 -3738 -6015 -9323 EBIT -322 -1510 -1908 -903 3757 Capital Raised 394.5 Secured Loan 250.0 -279.4 Cash Position 322.3 -1187.5 -3375.2 -4277.8 -827.2 Headcount 7 18 30 40 57 Avg Rev / Employee 34.3 50.3 61.0 127.8 229.5 1 EOY 2018 numbers based on 6 months to Dec EFTA00803403  Funding Seeking a £5m funding round to implement initial strategy PEOPLE PARTNERSHIPS PRODUCT Management / advisor to Distribution partners to help scale A more feature-rich platform supplement business experience the sales effort. including continued development of our prediction market service Sales team core to driving business Workforce provider to answer the expansion, including in the US. initial "who's going to do it- Focus on ease of use to avoid question and allow us to provide initial "consultancy burden" on our Data scientists to address an end-to-end service if required data scientists "consultancy burden" and build / support value-add services. Data aggregation platform to Cloud security requirements for provide distributional services to large sell-side organisations Developers to expand the platform new "alternative" data vendors EFTA00803404