Sana Labs

US-wide worker and income instability predicated on race and gender inequality drive the marginalisation of vulnerable communities. At the root of the problem is a vast nationwide skills mismatch, with over 7 million jobs left unfilled in 2019. The solution is dynamic, continuous - lifelong - learning. Yet, learning is inefficient, meaning learners lack engagement, and mastery, if reached, is rarely maintained. Ranked best-in-world for identifying knowledge gaps, Sana’s AI-powered learning platform provides optimal content at the optimal time to maintain engagement and boost knowledge retention. Sana ensures each learner: reaches proficiency at their own pace, masters the same content in half the time and remembers it three times longer. At scale, Sana democratises access to the skills workers need to keep up, transition, or go further. From boosting compliance for global pharma to reskilling nurses during the COVID-19 pandemic, Sana optimises the world’s most valuable resource - human capital.

Over 7 million workers across the US are poorly equipped to keep up with the changing job market, with many continuing to learn skills misaligned with the economy’s fastest growing sectors. By 2022, 27% of roles are forecast to be in jobs which don’t yet exist, and skills are becoming obsolete at an increasing rate, in many cases in less than five years. 

At the root of the problem are three core drivers: opaque skill demand; poor learning standards; and inequalities in access. Too often, industry needs are poorly defined, and even when they are clearly communicated, time and resource constraints mean professional learning is neglected and inefficient. Traditionally low engagement results in slow mastery and prolonged courses result in poor knowledge retention. Compounding these effects, access to upskilling and reskilling is reserved for the privileged few, meaning that entrenched race and gender biases broaden the gap between those empowered to keep up and those condemned to obsolescence. 

We believe our human-centric approach to workforce learning which: increases transparency between employers and employees in the market for skills; provides exceptional personalized and engaging learning; and is readily accessible to all, on demand, is the key to combating this skills mismatch. 

We are solving for universal access to world-class, personalized reskilling and upskilling. 

Sana Labs’ world-class adaptive learning platform provides a user-centric experience for learners and powerful authoring and analytics capabilities for employers. Not only does Sana empower workers to engage in lifelong learning, ensuring long-term skills liquidity, but it provides detailed analytics linking learning and business outcomes, promoting effective employer-employee coordination of skill requirements.

Learners undergo a three-stage journey to mastery: adaptive assessments which pinpoint current knowledge in up to 75% fewer questions than benchmarks; personalized learning paths which continuously assess proficiency, resulting in 15-45% greater learning outcomes; and targeted retention exercises, improving long-term retention by 3-4x. 

Administrators are empowered in real-time with: actionable insights and rich learning analytics, allowing them to engage personally with learners; grouping of users into roles and courses into paths, automating course assignment and improving follow-up; and a powerful authoring tool for data-driven content improvement. 

Sana is powered by state-of-the-art AI models which combine learning (including knowledge, remediation, repetition and progression) and content analytics (including content dependencies, difficulty, assessment and instructional value) with learning objective-aligned ‘nudges’ and data-driven content improvement recommendations. In doing so, Sana creates powerful, engaging, and fully personalized learning experiences at scale. 

Sana serves workers in sectors ranging from healthcare and pharma to financial, industrial and professional services. Represented by the diversity of our clients and partners, we have developed a deep understanding of the needs of the US workforce. We engage continuously with employers and employees alike, both deploying our end-to-end learning solution, and helping them author courses directly. Nurturing long term relationships, we refine our solutions to meet learners’ needs, whether integrating course ‘nudges’ to promote self-led learning, aligning dashboard metrics with client KPIs or integrating automated course pathways for minimal supervision. 

It has long been recognised that there is a need to shift from “mass standardization” to “mass uniqueness” to address the needs of today’s workforce. We envision a world of truly lifelong learners capable of mastering new disciplines, transitioning careers and augmenting their daily lives through intuitive and stimulating learning. Whether mastering new skills for a promotion or leveraging existing skills to break into new sectors, cutting-edge personalization algorithms ensure our learners stay engaged, reach mastery sooner, and retain it for longer. Fundamentally, our affordable, accessible and scalable solution empowers the “workers” of today to become the “shapers” of tomorrow, regardless of income, culture or background.

The problem, a US-wide skills mismatch worst affecting underrepresented citizens and driven by: opaque skill demand; poor learning standards; and inequalities in learning access, is closely aligned with the challenge at hand. 

By offering a scalable, low-cost, and universal platform designed to empower anyone to become a lifelong learner, Sana directly addresses these drivers to ensure that the American workforce keeps pace with its rapidly evolving economy. Sana not only serves as a facilitator for lifelong learning, but increases access to quality, affordable upskilling and reskilling opportunities and, through in-built course impact analytics, improves alignment between education and industry needs. 

Our team consists of: 22 Full-time employees.

At Sana Labs, inclusion and diversity are central to who we are and what we build. As a culture of brilliant innovators, creatives and dissenters, every step of our hiring is defined by an individuals’ potential to flourish with us, and each promotion is wholly meritocratic. We know diverse teams are stronger, more innovative, and more resilient. We actively hire people with diverse backgrounds, education profiles, nationalities and ethnicities, and judge their potential as an individual and as a prospective member of our team.

Yet, we recognise that even the most diverse companies can manifest unconscious biases. To us, a truly inclusive organisation is defined not only by the diversity of its workforce, but by the universal value-creating role of each service it offers, regardless of the end-user. As such, we carefully consider in-built biases and constantly revise our approach to ensure that the value we create is truly universal.

Our solution is innovative as it brings an entirely content agnostic, unbiased outcome-based approach to learning. 

Similar solutions to ours primarily measure success on seat-time (for example, how much time someone has spent on a particular module), or they measure a learner's self-confidence in knowledge acquired (most commonly by surveying learners after completing a module on how much they've learned). However, our solution is unique in that it applies recent breakthroughs in AI and reinforcement learning to objectively measure how much a learner knows, create an unbiased and adaptive learning curriculum based on their unique knowledge gaps, and model their knowledge decay based on content difficulty and historical knowledge retention.

Our core technology is built on Bayesian IRT models coupled with a rules-based engine to create an artificially-intelligent learning experience. 

Sana's core components include adaptive assessments, which accurately pinpoint knowledge gaps for individual learners; adaptive learning paths, which deliver personalized learning curriculums in real-time; and knowledge retention models, which estimate the half-life of the acquired knowledge for a specific learner.

In addition to the adaptive learning experience, we also support browse and search, with the search powered by a deep learning algorithm that surfaces results based on imprecise search terms, badly indexed data, and retrieves courses with minimal metadata.

These core components are built on a world-class infrastructure capable of supporting millions of learners simultaneously, offering personalization and personalized learning analytics in real-time.

• Bayesian IRT models are used worldwide in high-stakes exams such as the GMAT and GRE. 

• Deep learning search algorithms are used extensively in web browsers world-wide.

• Modelling the half-life of an individual's knowledge is a technique that builds on the work of optimizing spaced repetition intervals researched by many learning science and cognitive science academics worldwide. 

In this video (already shared above), we have customer testimonials of how and why the technology works.

A changing world necessitates a changing workforce. The skills mismatch affects 2 out of 5 employees in OECD countries and creates trillions in global productivity losses each year. BCG research demonstrates that, pre-COVID this mismatch effectively imposed a 6% annual “tax” on the global economy in the form of unrealised productivity. While the worsening global skills mismatch has long been recognised as increasingly detrimental for economic performance, the COVID-19 pandemic has worsened these associated productivity losses from 6% to 11%, meaning over $18 trillion in unrealised global GDP by 2025.  Such a significant acceleration in global demand for effective upskilling and reskilling of workers across industries only reinforces the need for effective workforce learning at scale. 

Yet, presently scalable solutions, such as traditional Learning Management Systems and generic online course offerings, are inefficient: workers lack motivation to learn, their engagement is poor, they reach mastery inefficiently and have poor knowledge retention if mastery is reached. While personalisation is the holy grail of learning, most companies cannot afford to invest in one-on-one training for their entire workforce, and so settle for poor substitutes with low return on investment. 

By harnessing AI and ML, we provide a scalable solution capable of addressing individual learner needs worldwide and in real-time. Having launched our solution with one of the largest and most prominent health systems in the world at the height of the COVID-19 pandemic, we have firmly demonstrated that our approach meets the US-economy’s pressing need for skill liquidity. In particular, generalist nurses using the Sana platform as part of Project Florence reached mastery as COVID-19 specialists twice as fast as by traditional linear approaches, and retained their knowledge up to the times longer. 

Learning is at the heart of reskilling. Given that our solution is entirely content agnostic, we are confident that we will be able to address the worsening skills mismatch across every sector and state over the coming year, limiting mass productivity losses in light of the COVID-19 pandemic to aid efficiency and long-term economic recovery. 

In the next year, we intend to broaden our learner base to 200,000 to consolidate the proven efficacy of our solution across a greater breadth of sectors and geographies, focussing in particular on meeting the growing and evolving needs of global healthcare. 

Our long-term impact goals are to expand the educational content on our adaptive learning platform from proficiency, upskilling and reskilling, to include other socially-relevant topics such as building awareness of LGBTQ+, Civil Rights, and cultural norms that improve labour market access, skills liquidity and mobility. In doing so, in the next five years we intend to reach not only healthcare and business professionals, but over 2 million learners across all industries and from all backgrounds who can also benefit from learning topics which can help them become better global citizens. 

We plan to partner with organizations who specialize in building awareness of these socially relevant topics, and look to leverage our adaptive learning platform to bring their educational content to scale.

Our biggest hurdle over the next five years will be ensuring knowledge and market openness to our unique approach. A broad lack of public understanding of AI as an unexplainable “black box” is likely to hinder adoption as we scale and thereby limit our scope for global impact, especially in the traditional (energy, industrial, manufacturing and agricultural) sectors positioned to benefit most from efficient upskilling and reskilling. 

In the next year, we face the challenge of increasing awareness of the benefits of AI, as well as its limitations. With this, we will need to proactively shape the hearts and minds of communities most in need of Sana, laying firm foundations for our widespread impact on the US workforce.

We plan to overcome these barriers by:

1. Providing free AI courses on our learning platform, both to educate users on the applications of AI and to demonstrate the efficacy of our learning solution. One course is already publicly available at sanalabs.com

2. Partnering with thought-leading organisations in as many sectors as possible, in order to build a diverse set of use-cases and contexts for our solution to demonstrate the benefits of AI in learning

3. Harness case studies, demos and thought-leadership to provide our customers with the insights required to understand the applications of AI and demonstrate their workers’ benefit from learning with us. 

SOLVE and various Challenge partner communities play a critical role in this regard, increasing awareness of the efficacy and safety of our solution, and providing much-needed credibility when seeking early partners for reskilling and upskilling within currently undisrupted industries across the United States. 

None.

N/A

Our team consists of exceptional individuals in every domain. Valentino, our Director of Business Development, Oscar, our Director of Research and Susanna, our Senior Data Scientist, have attained PhD’s in the fields of Computational Linguistics, Electrical Engineering, and Applied Mathematics, respectively. Valentino was not only the first in his family to attain a PhD, but the first to attend university altogether. Prior to joining us at Sana, they worked on Google AI Search optimization; built powerful analytics-based solutions with BCG GAMMA; and harnessed predictive modelling to optimise drug therapies with AstraZeneca. As Google Scholars, they have produced a plethora of research papers in the fields of computational linguistics and natural language processing while at Sana. 

Those traditionally underrepresented in upwardly mobile careers are those whose opportunities have been restricted by family circumstance, race, gender, or sexual orientation. As a diverse team, we are all too familiar with the challenges of economic inequality and discrimination in preventing equal access. Raised in over 11 countries, we represent a broad spectrum of cultural perspectives, economies and customs. Each of us has experienced the ‘leap of faith’ of reskilling first-hand as we navigated our own tough life transitions, in many cases coming from backgrounds where the odds were stacked against us. We believe the unique and exceptional life trajectory of each member of our team makes us well-equipped to empathise closely with our end-user - namely, people seeking better lives: whether that means improved job security; income stability; or simply personal satisfaction and happiness.

We are partnering with Mount Sinai Health Systems in NYC, King Faisal Hospital in Saudi Arabia, and Karolinska Institute in Sweden, to upskill nurses in optimal treatment for COVID-19.

We are also working together with smaller hospitals across Europe, the Middle East, Asia, Oceania, and South America to deliver specialist training in response to the recent COVID-19 pandemic.

In addition to this, we partner with some of the world's largest Pharma, financial and education services companies.

We provide our customers with an SaaS license-based platform product which creates an on-demand adaptive learning experience, provides analytics on learner performance and content richness, and enables administrators to author content in an adaptive-integrated method.

We also provide services which include the integration of existing content into our adaptive learning platform, and the integration of our adaptive learning system into other systems in the customer's ecosystem.

Customers ask for our product and services because they want to transform their learning experience to achieve better learning outcomes, faster. Every aspect of our service is designed to deliver on this: from our intuitive and rich user experience to seamlessly integrated cutting-edge AI to model learner knowledge, knowledge gaps, and knowledge decay. Specifically, learners undergo a three-stage journey to mastery: adaptive assessments which pinpoint their current knowledge and confidence in up to 75% fewer questions than benchmarks; individually personalized learning paths which continuously assess proficiency, resulting in 15-45% greater learning outcomes and a 61% reduction in time to mastery; and targeted retention exercises, improving long-term retention by 3x.

In doing so, Sana delivers a superb learning experience, demonstrated by an Net Promoter Score (NPS) of 74, which sharply contrasts the NPS score of -25 for Learning Management Systems today. This is a testament not only to the fact that Sana empowers users to achieve their objectives, but to Sana’s evident usability, efficacy and impact on learners and their organisations. 

We will fund our work primarily through revenue, selling product licenses on a SaaS basis. Further, as we seek to scale our solution even more rapidly to meet the demands of an increasingly rapidly changing workforce, we may also raise more investment capital over time.

Since it was founded in 2016, Sana Labs has raised $2.5 million from some of Europe's most prominent private investors and received grants of $100k from Vinnova and has been generating revenue since 2018.

We do not state this publicly.

We do not state this publicly.

Since our greatest barrier is awareness of and knowledge about AI-centric approaches to learning, we are applying to SOLVE to gain credibility for our unique approach and raise awareness of the power of AI in workforce learning. 

Indeed, while the Solve prize money will inevitably help us invest in scaling and improving our technical capacity through R&D, the Solve network itself promises substantial opportunities both for revenue growth and innovation aligned with our long-term goals.

In particular, becoming a Solve team offers the opportunity for cross-discipline collaboration with Solve alumni, across a breadth of complementary domains including: community-driven innovation, teaching & education, frontlines of health and work/workplaces of the future. By sharing intellectual and human capital with fellow "Solvers" and the global MIT network, we intend to realise the full societal scope of personalized learning as a tool for citizen inclusion and integration, as well as professional growth and economic development. This not only enables us to serve a broader scope of clients, and thereby pursue diversified revenue streams which help long-term financial profitability and sustainability, but validates our mission and efficacy. 

Thus, becoming a Solve team is likely not only to directly alleviate our financial capital constraints, but energise and augment our existing R&D efforts to better serve our existing clients and acquire new ones across new domains, and increase both our brand awareness and commercial exposure. 

Our primary partnership goals are, in the short-term, healthcare, life sciences, and professional, industrial and financial services clients seeking to ensure proficiency, skill improvements and skill diversification. This is a substantial global market which has long employed outdated "one-size-fits-all" online learning platforms which are inefficient and result in poor engagement and utilisation. This provides a substantial opportunity for our personalized approach to drive significant learning gains and generate invaluable case studies required for our solution to gain traction among communities traditionally less open to advanced technologies such as AI. 

In the long-term, we seek to partner with public sector companies, governments and nonprofits to roll-out personalized learning technologies as tools for civic engagement and economic development, especially supporting vulnerable and marginalised communities and citizens. 

The initiatives we’d like to partner with are: UNHCR, The Bill & Melinda Gates Foundation, the Chan Zuckerberg Initiative (CZI) and Teach For All, each of which have personalized learning at scale as one of their key focus areas. A partnership seems natural and would give Sana Labs additional power to be part of a worldwide learning solution alongside established players in development and emerging market innovation.

The organizations we’d like to partner with are Fortune 500 companies that experience a strong need to upskill and reskill their workforce, such as but not limited to General Motors, Walmart and Verizon. With the Sana Platform, they can ensure their employees not only thrive in their existing roles, but are capable of developing rapidly to occupy new essential roles within the organization, rather than having to lay employees off due as their skill sets become outdated. Consequently, we believe the  Sana platform is well positioned to ensure long-term skill liquidity: the most vulnerable low-income individuals dependent on these organizations for survival being empowered to grow their skill set and stay employed.

Naturally we also aim to partner with more pharma and healthcare organizations like Johnson & Johnson, Novartis, United Health Group and CVS Health. As stated previously we have proven the impact of our solution in this category and envision fruitful partnerships in multiple domains.