UTU – Human Centered Trust for the Digital Economy

We believe in a more human-friendly internet, accessible to all. We are pioneering digital models of trust built around how humans naturally build confidence in one another. UTU’s vision is to become the trust infrastructure of the entire internet, replacing anonymous star ratings, reviews, and scores. Our mission is to protect privacy as we make the internet a safer, more trusted place to gather, share, work, and trade. 

We do this through leveraging of our taxi/mobility app franchises across Africa, for which our trust AI was invented. As we’ve grown from one to five countries in 2018 and aim to reach another 10 in 2019, our franchise model leverages local entrepreneurs and knowledge decentralize tech ownership.  

Our success means both massive, decentralized wealth creation for our franchisees across Africa and better positions ourselves to provide access to trusted tech-enabled home, family, and financial services across Africa.

African and indeed human markets function based on networks of trust, yet this type social trust is absent from digital marketplaces, constraining growth of the gig economy, which can more efficiently deliver services to grass-roots consumers, grow the livelihoods of informal workers, and transition them to the formal economy. Instead, digital trust is based in poor proxies such as aggregated, anonymous star ratings, reviews, and scores, all of which inadequate measures of trust for the majority of Africans to trust these services. We tap into traditional models of trust and digitize them to overcome this challenge. 

Furthermore, centralized ownership of tech platforms limits the reach of these benefits only to the largest African cities, leaving the vast majority unserved and ultimately doing little create wealth at local levels. By tackling both trust and ownership models we unlock the entrepreneurial potential of the continent to bring trust, tech-enabled services to the more than 60% of Africans.

We provide two distinct B2B2C services. Together they support more than 11 million people’s access to mobility, financial, and other tech-enabled services. 

1.     For mobility, we provide franchising as a service to local entrepreneurs for no money up front to bring a world-class mobility app to their market. We are working with entrepreneurs in 20 countries across Africa to launch mobility franchises in their markets. Our goal is to launch in the 500 largest African cities by end of 2021, providing more than 600 million people access to trusted mobility services.

2.    We have successfully demonstrated the value of our trust infrastructure for multiple other sectors of the gig and P2P economy, by digitizing the existing real-world models of trust on which these industries operate and serving it up to these marketplace platforms as the trust infrastructure as a service (via API). Our clients are platforms for peer to peer financial services, mobility, local on demand services, hiring/talent acquisition, telehealth, and eCommerce. Our flagship client is a P2P lending platform in Tanzania with which we are currently digitizing the records of 11 million small farmers and building trust models for them to help them access financial services.

UTU has two solutions that mutually reinforce and support each other. 

• Mobility App Franchising Platform – We started the first taxi app in Africa back in 2013, based in Nairobi, Kenya. Without big funding to compete against the onslaught international taxi apps, we innovated both on product and business model. We recognized that taxi apps from the west mistook taxis for commodities, not individuals that have direct relationships with their clients, built on mutual trust. We reconceptualized the taxi app from the ground up to focus on interpersonal trust, the result being our trust AI. We then opened up our platform to let entrepreneurs across Africa bring it to their market with limited investment required. By centralizing development costs and decentralizing go-to-market through empowerment of local entrepreneurs to make money by bringing safe transport solutions to their town or city, we have unlocked the ability of tech-enabled services to reach the majority of the African population, rather than the current 5%. 
• Trust Infrastructure as a Service (for all sectors) – Recognizing massive benefits to our clients from our Trust AI, we extracted it from the mobility application to deliver AI-powered Trust Infrastructure as a Service though APIs, SDKs, smart contracts, and oracles to help sharing platforms and digital marketplaces build better trust across every sector, particularly any transactions that expose a user’s family, home, health, assets, or business to any type of risk. Our trust engine drives all key marketplace metrics including acquisition, conversion, and satisfaction by building individualized, contextually aware trust models for each user based on their relationships, in-app behaviors, and other contextual factors. We are also tokenizing the building of trust through our blockchain protocol, to provide economic incentives for building trust.

To help our franchises scale, we structure costs according to their needs, offer hands-on training and on-going advisory services, content and systems for marketing, operations, business intelligence, finance, and fundraising support. In the future we hope to provide a franchisee development fund to provide matching capital for locally raised funds. 

UTU has generated significant IP, including a Utility Model from the Kenyan Intellectual Property Institute and a pending US/PCT Patent. We are currently pioneering novel ML techniques for learning from sparse data and graph embedding. We have published our blockchain protocol here: https://docs.google.com/document/d/1UFFgZL2_Krc-cgAm_mM1eNJdABF1bomdiXoh2RZVIsk/edit#heading=h.871wg7mh1vjj,which we have designed to eliminate ratings/review fraud, incentivize trust-building behaviors, and empower user ownership and control of their data.

We have several key innovations:

• Theory of Trust - Humanized, personalized, contextualized. By evaluating trust first and foremost based on the presence/strength of our users/providers’ relationships, evaluating buyers and sellers in context of one another rather than in absolute terms, and adjusting for sector-relevant environmental factors, our trust engine comes closer to modeling human trust than any competitor.
• Tech Franchising Model – While most tech companies seek to raise massive VC funding to push their solutions, this neither works well for Africa’s highly distributed population nor creates local wealth nor capitalizes on Africa’s biggest competitive advantage, entrepreneurs. Our model does all three.
• Data Acquisition Channel - Our mobility franchises serve as a critical user/data acquisition channel that we use to train our AI. Mobility is an ideal data catchment as an easily adopted tech-enabled service that can often serve as a gateway to using other app-based service. The user-base generated in each of our mobility franchises opens the door to our diversification of trust services to other sectors in those geographies, creating a virtuous cycle of new opportunities for franchisees and gig workers on the platform. 
• Market Position – As a neutral third party provider of trust infrastructure, rather than trust mechanisms built by marketplaces directly, we become the “honest broker,” transforming trust from a product into a currency. We operationalize this by offering our trust services as Infrastructure as a Service.  

On the highest level, this system is a composition of two main components:

1. Recommendation service using AI/Machine Learning: serving provider recommendations for requested services, it is currently based on a sparse dataset of transactions. For this purpose, we encode current and historical request contexts into feature sets and use a matrix computation to identify their similarities. E.g. for the taxi application, whether are ride happens during a weekday or weekend, in the morning, afternoon or night etc. Thereby the users most similar to the current requester and their most preferred providers for similar requests are identified. The best recommenders are then chosen for display based on this as well as their connection and common relationships in the social graph. The defining features that lead to the similarity are then extracted and displayed in a user-understandable way.

2. Decentralized and blockchain-based protocol to facilitate data provision: UTU protocol is being implemented as a suite of smart contracts and oracles on the æternity blockchain, which uses Bitcoin-NG and State Channels for better, cheaper scalability. UTU Coin will be implemented similarly to an ERC20 token, whereas UTU Token will be implemented as a non-transferrable token. Smart contracts let clients define access rights and fees to their data, while data itself is stored off-chain at locations of the client’s choice. Each of the contracts will be made available in privacy-preserving variants to allow the restriction of the visibility of any data, as well as allowing for anonymization. 

With six years’ experience building and growing a sharing platform across Africa, we have come to understand the core challenges facing not just our platform but other sharing platforms as well. UTU was originally created to solve for our mobility app platform but has broad-based applications to revolutionize the sharing economy.

Challenge 1 – Capital requirements. While many platforms’ technology can be built for relatively little money, the go-to-market phase can be extremely expensive, especially in Africa, where population is highly distributed and purchasing power is low. Our franchise model, which can be replicated in other sectors, turns expansion from a cost center to a revenue center, and let’s franchisees become profitable with low volumes since they have no engineering team to support, while our fees are tied to revenue.

Challenge 2 – Localization vs. scalability. Scaling products across the 52 unique and distinct countries of Africa, each with unique submarkets, can cause many products to fail. Letting local entrepreneurs, who understand their markets, drive product development leads to a highly localized product for each market but with the benefits of an international footprint. 

Challenge 3 – The lack of digital trust in societies where trust is primarily interpersonal rather than institutional, massively constrains uptake. By bringing real world trust mechanisms to digital transactions, we believe we can help platforms reach a majority of their target markets. 

We currently serve more than 50,000 people in Nairobi alone with growing user bases in our new markets. Through our trust infrastructure service, we are supporting 11 million farmers in Tanzania to get better access to financial services through our partner Jamborow. By the end of 2020 we expect our mobility business to be on the ground in 250 cities across the continent, serving more than two million people monthly. By the same time we believe our trust engine could be providing services to platforms that collectively serve more than 100 million people around the world. In five years, we believe that our trust infrastructure services will be the global market leader for trust infrastructure as a service, the category we are pioneering, as common a web-based tool as AWS, Salesforce, or PayPal.

Our goal for the 18 months is to launch our mobility platform in 250 cities across Africa through a network of local entrepreneurs. Within five years we want to be present in every town over 50,000 people on the continent where we can transform mobility access at the continental scale. With our trust infrastructure as a service, we seek to become the global standard for trust infrastructure, a category we’ve invented, replacing anonymous star ratings and reviews, as well as “scores” as the de facto trust mechanisms of the internet. Our mission is to protect privacy as we make the internet a safer, more trusted place to gather, share, work, and trade.

We face three key types of challenges:

Technology - There are many technical challenges to overcome with our product, given the cutting-edge R&D work our team is doing to effectively digitize and tokenize real world trust. These challenges relate to machine learning from sparse data, digitization of physical information, modeling of contextual factors into personalized trust models, and the enforcement of privacy mechanisms such as k-anonymity.

Interface Design – Human models of trust can be very difficult to quickly and meaningfully communicate through digital means. They require highly intuitive designs and extremely precise use of generated language. Designing new user experiences that simplify the process of understanding trust in these contexts will be among our chief obstacles. 

Market Barriers – Ultimately our ability to provide trust infrastructure to the world lay in our ability to train our AI through our mobility application. Thus the scaling of our mobility business is critical. The key challenge here will be our ability to withstand the massive budgets and predatory practices of international ridesharing giants the ultimately end up being highly extractive. The corollary being our franchising model which can lead to variances in quality of service and experience across markets if systems are not well designed and implemented.

Technical – Despite the scope of the technical challenges at hand, we have the most capable AI team in Kenya. Our CTO did his PhD on distributed AI for trust and our Head of Data Science has a PhD in High Energy Nuclear Physics and has built AI models on the largest particle collision datasets in the world. We also have an R&D partnership with the Agents, Interactions, and Complexity Department of the University of Southampton in the UK, where we are collaborating with various faculty and post-docs to solve these challenges.

Interface – Our mobility application, especially the Nairobi Franchise, which we operate ourselves, has already proven to be the key learning ground for us. Unlike in other sectors where we serve up our trust infrastructure via API, in mobility we run the entire product down to the end user. This allows us to experiment with different interface mechanisms (app, SMS, USSD, WhatsApp, voice) as well as different design schemes and ordering protocols that help us remove biases from our data collection.

Market – We have avoided getting into pricing wars with our global competitors, refused to subsidize rides, charge exploitative commissions, or engage in anti-competitive practices such as debt-bondage via car loan that larger mobility apps use to establish power over their driver network. Instead we’ve focused on creating value for drivers, helping them grow their businesses, and operate more efficiently. To this date we are the only large taxi app in Africa that has never suffered a driver strike. 

Our team is 15 full time staff and 10 interns based in Nairobi, Kenya. Our staff comprises 8 engineers and 7 operations, sales, and customer service professionals. Our interns split similarly. Our network of franchisees comprise another 15 people pushing our platform in their markets.

Our leadership comes from the Kenya (2), the US (1), and Germany (2) with 70 years collective experience across strategy, operations, AI/ML, computer science, and finance. Our mobility platform, MARAMOJA, already scaling across Africa gives us unique perspective on the challenges of growing sharing marketplaces in emerging markets,  a strong revenue stream and a proprietary data acquisition channel. We also have an R&D agreement in place with the top University Department for AI in the UK.

Before founding UTU, CEO Jason led strategy, design, and development of large-scale donor projects across multiple sectors in more than 40 countries, helping secure more than $3B in funding opportunities. CTO Dr. Bastian Blankenburg, focused his PhD on distributed AI utilizing multi-agent systems, the foundation of our IP, and previously built and deployed large-scale transportation systems throughout Europe and the Middle East for a leading German transportation software company. Director of Mobility, Ronald Mahondo brings more than 12 years in the African private sector, including senior leadership roles in the technology, banking, and oil & gas sectors. Before joining UTU, he led two other major mobility brands from 0 to profitability in Africa and managed annual budgets of up to $10M CAPEX and $100M OPEX. Head of Data Science, Dr. Alex Mwai has a PhD in High Energy Nuclear Physics and has built AI models on the largest particle collision datasets in the world.

MARAMOJA / FLEX / COCO Ride / CHE – Scaling UTU’s footprint and training our AI through mobility sector franchises across Africa.

Æternity blockchain – Funding and blockchain support, our protocol is being built atop their blockchain

SoftBank Group Subsidiary DEEPCORE and Zeroth.AI – Funding and AI support

University of Southampton – Agents, Interactions, and Complexity Group – R&D partner

Syntax Ltd – Adapting our core mobility technology to serve goods transport/logistics use cases

Jamborow – Implementation of our trust infrastructure for P2P lending with 11 million users in Tanzania

Graychain – Crypto credit bureau with whom UTU has a data sharing partnership to enrich each others’ datasets

Netwookie – Implementation of our trust infrastructure in Kenya for local on-demand services including beauty, wellness, home, and family services

RideSafe – Implementation of our trust engine for motorcycle taxis to access emergency roadside services

TipMe – Implementation of our trust engine to provide global direct tipping to factory workers

Wiseper – Implementation of our trust engine to mitigate the global fake news challenge

We have two business models: 

1.    Our mobility franchises around Africa pay licensing fees as monthly revenue shares. We are currently on the ground in 10 cities across 5 countries with franchise agreements already in place for 10 more African countries, with another 5 countries under negotiation, mostly in markets where we are or will be the only mobility app on available. We currently serve more than 50,000 passengers and drivers across our network. Our franchise model effectively enables small entrepreneurs to bring world class mobility services to their local towns with a world class product, tailored to the needs of their market. We take a minority revenue shares and equity stakes to enable franchises to reach profitability quickly with a path to scale. 

2.    Our Trust Engine builds on the learning of the mobility franchises to provide Trust Infrastructure as a Service, charged monthly either based on consumption (# of queries to API) or success (customer acquisitions, conversions) based on client preference. Price per query varies in the low single digits and fractions of cents based on client volumes. Success fees are calibrated to yield similar slightly higher revenues (on a per query basis) to account for extra risk taken by UTU in only charging for success. Collectively our platform clients serve more than 11 million users across Africa.

We have not taken any grants or donations, only commercial debt and equity investments. Our revenues from the mobility business are growing substantially that we could be operationally profitable at the group level by mid 2020. Mobility revenue and equity investments in turn together finance the development of our trust AI, which has significant R&D costs and a longer path to profitability which we expect by 2021.

We see a number of benefits to being a Solver:

1. We are pioneering new AI/ML techniques across distributed multi-agent systems and would love access to MIT's network of academics and experts in the field.

2. Many Solvers seem to be running platforms and marketplaces that we believe could benefit from building better trust among their their users that would also help us demonstrate our value proposition in new sectors and geographies.

3. As a small African startup, it can be hard to break out of the "Africa company" narrative even if you have a globally relevant value proposition. We believe MIT's credibility, networks, and partners could help us cross that chasm.

We are open to work with any digital two-sided marketplace across the range of sharing economy sectors as well as any other large corporate with a digital footprint whose growth or satisfaction is constrained by barriers relating to trust.

We believe that societal trust is being eroded by the imposition of inadequate digital replacements that are unrelated to human trust, highly prescriptive, biased, and lacking nuance. Moreover, authoritarian regimes use of citizen scorecards further erode interpersonal trust and invade privacy. UTU is passionately developing alternative AI models of digital trust that put humans first, describe trust in human terms on a context-aware individualized basis rather than prescribing it against a fixed universalized and static scorecard, and preserve user ownership and control of data and privacy in the process. We face a number of significant AI challenges for which we are trying to build our networks and access to top tier talent as much as possible, most of whom are beyond our reach as a small, Kenyan startup.

We believe that societal trust is being eroded by the imposition of inadequate digital replacements that are unrelated to human trust, highly prescriptive, biased, and lacking nuance. Moreover, authoritarian regimes use of citizen scorecards further erode interpersonal trust and invade privacy. UTU is passionately developing alternative AI models of digital trust that put humans first, describe trust in human terms on a context-aware individualized basis rather than prescribing it against a fixed universalized and static scorecard, and preserve user ownership and control of data and privacy in the process. We face a number of significant AI challenges for which we are trying to build our networks and access to top tier talent as much as possible, most of whom are beyond our reach as a small, Kenyan startup.