Upski: Smart Bot for Economic Mobility

Pat Pataranutaporn is an antidisciplinary technologist/scientist/artist at the Massachusetts Institute of Technology (MIT). He is part of the Fluid Interfaces research group at MIT Media Lab led by Professor Pattie Maes, which specializes in designing on-body technology for human enhancement. Pat's research is at the intersection of biotechnology and wearable computing, specifically at the interface of biological and digital systems.

Pat has worked with global collaborators such as NASA TRISH, IBM Research, Bose, Harvard, UCSB, ASU, NTU, and the Futuristic Research Cluster of Thailand (FREAK Lab) to examine the symbiotic relationships between human and technology. His interdisciplinary research ranges from investigating human-AI interactions, developing wearable labs on the body with programmable bio-digital organs for space exploration, developing machine learning models to detect linguistic markers related to mental health issues, and developing and designing mind-controlled 3D printers.

Beyond the health crisis brought by COVID-19, it has also led to global economic disruption that disproportionately affects certain populations: people in less-protected and low-paying jobs, especially the youth and elderly; women; and migrant workers. This will further increase global socio-economic inequality.

Upski mitigates this damage by using technology to help connect job seekers with remote work opportunities as an alternative income source. It is an intelligent recommendation system (chatbot) that introduces remote job opportunities to our beneficiaries. We assess their skills and interests and match them with appropriate remote work and upskilling resources.

Beyond the immediate impacts of remote job matching during the COVID-19 crisis, in the long term, we aim to end poverty and inequality for everyone. Our broader vision is to improve the economic mobility of the most vulnerable in the workforce by providing equitable access to quality upskilling and remote job opportunities.

The COVID-19 pandemic has led to global economic disruption. There will be close to 200 million job losses globally, and 26 million people in the U.S. alone have already applied for unemployment benefits. Economies worldwide are expected to experience substantial contraction -- it was predicted that the 19 nations using the euro would contract by 15% in the first quarter of 2020 and 22% in the second. 

Additionally, certain populations are disproportionately affected by this job crisis: people in less-protected and low-paying jobs, especially the youth and elderly; women; and migrant workers. This will further increase global socio-economic inequality. 

Research shows that job loss is linked to increased depression, anxiety, and stress, and subsequently higher rates of suicide. People experiencing job or income loss during COVID-19 have reported higher rates of negative mental health impacts compared to those not experiencing job or income loss. Therefore, these populations are experiencing a double health risk with COVID-19: the actual physical health effects if infected, as well as the mental health effects brought on by the financial stresses of losing a job as a result of the pandemic. The already vulnerable are becoming more vulnerable. 

Our team mitigates the socio-economic damage caused by COVID-19 by using technology to help bridge the gap between businesses needing to recruit for remote positions but also needing to recover from financial losses caused by COVID-19, and job seekers from marginalized communities whose dire living situations are worsened by COVID-19. 

Upski is an intelligent job-matching system in the form of a chatbot. Our business-facing chatbot determines open job requirements and company culture, while our people-facing chatbot assesses job seekers’ skills and interests. Upski then cross-checks the data between the two chatbots to recommend the best matches for both the employers and potential employees. Building this access bridge for our target demographic opens a range of possibilities to improve their economic mobility while also expanding the global talent pool reach for employers.

We have found that our target population is more open and eager to interact with our friendly chatbot in discussing issues like job security that typically require a level of trust, comfort, and security. Our platform is easy and straightforward to use and does not require much technical knowledge. Additionally, unlike most available alternative job platforms, our chatbot provides tailored job matches based on individual skills and interest assessments.

Our target populations are those most impacted by the COVID-19 economic downturn: workers in low-paying jobs that are getting laid off in higher numbers. These individuals are more likely to be the youth, elderly, women, and migrant workers.

Our Upski chatbot has reached thousands of vulnerable workers in 22 countries, with the Philippines accounting for over half of our users. This doesn’t come as a surprise since there are 2.5 million (migrant) Overseas Filipino Workers. The average age of all our users is 30 and 70% are female. These numbers align with our target population of those most affected by the COVID-19 economic downturn--youth, women, and migrants. Most significantly, 80% of respondents state that our job and upskilling recommendations include a resource they plan to pursue.

We combine user interviews with advanced technology to gain a deep understanding of and insights into the problem and our target demographic. Our multi-faceted approach of developing personal relationships while also developing the technology behind the service creates a grassroots appeal that is backed by data. As we continue to make our digital tool more intelligent, automated, and comprehensive, we can establish that bridge to sustainable economic mobility for those most in need.

COVID-19 disproportionately affects marginalized communities. The likelihood of job loss during COVID-19 is linked to socio-economic status, with the most vulnerable also the ones most likely to be laid off. Additionally, COVID-19 has pushed many companies to invest in technologies to automate jobs that have traditionally been held by people with lower-level skills. As a result, the skills gap in jobs is expected to expand after COVID-19, with fewer jobs available for those possessing lower-level skills.

Upski elevates opportunities especially for these people by bridging this gap in employment opportunities through equal access and upskilling for the future. 

Before the team launched Upski, we learned lessons from the COVID Bot, another chatbot developed at FREAK Lab in March 2020 that helps screen users for their risk level of contracting COVID-19. We pivoted to focus our recommendation system from healthcare, which was the more immediate primary need at the time, to job matching, which has become an immediate secondary need that also plays a crucial role as a long-term need in the recovery from COVID-19. 

Issues like health and job loss require a level of trust, comfort, and security to discuss. We discovered that by using a chatbot, people were more open and eager to interact with a friendly virtual assistant. We discovered gaps in the current job search landscape: lack of equitable access to information and reliable opportunities. User insights led us to include upskilling resources in one of our iterations. Furthermore, our next phase will be focused on employers to ensure time and cost efficiency in the recruitment process.

Our team is composed of Harvard, MIT, and Stanford graduate students, alumni, and tech entrepreneurs who are united by the passion to ensure the well-being and economic mobility of the most vulnerable in society. We all grew up in developing countries, with some of us even living through and experiencing the suffering and inequalities firsthand, specifically in the Philippines and Thailand (which also happen to be two of our key countries of focus with our Upski efforts). All of us have worked significantly in the issues in varying methods and capacities. Though we all have different expertise, what brought us together is our concentration within our own industries on addressing socio-economic injustices and working on bringing about global equity. We have narrowed down on the upskilling/education and employment aspects because of our own stories. We are academics and practitioners who have used education and our careers to be able to effect large-scale change, with some of us even using them to be able to break the cycle of poverty we were born into. 

COVID-19 ignited our team’s drive to get a product out to the public that would be helpful through the pandemic, and hopefully for even longer after.

The Upski team has the ideal combination of technical, design, and entrepreneurial talent backed by firsthand experience with and connections to the problem Upski addresses.

Our team is composed of the following core members:

• Pat Pataranutaporn is a big data and AI expert pursuing his PhD at MIT Media Lab. His work has been featured in notable publications like Time Magazine and National Geographic. He has won numerous social entrepreneurship and innovation awards, and he works with a global network of collaborators.

• Viria Vichit-Vadakan is a social entrepreneur recognized by the Obama Foundation and World Economic Forum for her work in education technology. She has worked at leading nonprofits, such as IDEO.org and Ashoka. She holds a BA in Urban Studies from Stanford and an MBA from Harvard Business School.

• Nikiya Crisostomo has a BS and MS in Civil Engineering from Stanford University, with her research focused on social impact through socially and culturally conscious engineering innovations. She has worked in engineering management, data analytics, and digital design. She grew up in poverty and migrated from the Philippines, so Upski’s mission is deeply personal for her.

• Poom Shetshotisak is an entrepreneur and engineer who worked in automation and robotics before he founded businesses that leverage chatbot technologies and digital marketing. As an entrepreneur, Poom understands how to build products that will best serve our beneficiaries at scale, and what employers need in terms of skills and talent.

When COVID-19 hit unexpectedly, the team had to quickly mobilize to launch a prototype. It was not within the planned timeline to launch that soon, but the economic impact caused by COVID-19 pushed us to act. Within a week, we had put together the chatbot to begin reaching out to individuals seeking remote jobs. Within the first five days of launching Upski in limited circles, we reached hundreds of vulnerable workers in 13 countries. 

Soon after overcoming the initial hump of developing a prototype for the public, the team was faced with another challenge. While the technical tool was delivering the basic service of recommending job platforms to users, the structure was not giving the team more in-depth information to truly understand the target demographic. The technical team scoured through the chatbot questions and responses and used the data to help improve the chatbot so that it could gather critical information. As a result, our user engagement rate increased by 27%. Additionally, the social team organized over a hundred one-on-one interviews with users to gather qualitative data as well as begin to build relationships to create the Upski network.

All of this was accomplished by the team without any funding. 

Our team members have demonstrated leadership in their respective communities to improve economic mobility and access to education. 

• In 2014, Viria founded a 4-year undergraduate BA program in Global Studies and Social Entrepreneurship in Thailand, providing access to quality education for more than 200 underserved students in Thailand, Cambodia, Myanmar, and Bhutan. At the same time, she was working with an education startup that was impacting more than 25,000 students in Mathematics and Science education. 

• From 2008-2012, Nikiya led the Stanford Society of Women Engineers’ largest event, annually inviting 200 elementary school students from low-income neighborhoods to experience Stanford through STEM workshops. More recently, she donated paintings and digital art for an auction to raise funds for Filipino workers who lost jobs due to COVID-19.

• Pat’s research has been globally recognized as creative innovator. Pat had worked with global collaborators such a NASA TRISH, IBM Research, Bose, Harvard University to examine the symbiotic relationships between human and technology. Pat's research has been featured on FastCompany, Time, Disruptive Innovation Festival, National Geographic, The Guardian, and UNEP. 

• Poom is a successful tech entrepreneur that has started companies that work in the area of environmental sustainability and health care.

-

Our approach combines user interviews with advanced technology to gain a deep understanding of and insights into the problem and our target demographic. We use supervised artificial intelligence to address socio-economic issues, targeting the most vulnerable in society that have typically had little to no access to remote job opportunities. Building this access bridge for these populations opens a range of possibilities to improve their economic mobility while also expanding the global talent pool reach for employers.

Our Upski chatbot is designed to be a friendly and approachable virtual assistant that helps understand job seekers’ unique needs and skills, and connects them to relevant remote work and skills learning resources using the artificial intelligence recommendation system. The recommended jobs have been pre-screened for credibility, making the job search process easier and more straightforward for our users. 

Additionally, this chatbot serves as the social probe for collecting socio-economic data needed for big data analytics and visualizations of the COVID-19 pandemic’s effects on the job market and global economy.

Our multi-faceted approach of developing personal relationships while also developing the technology behind the service creates a grassroots appeal that delivers sustainable high-quality results backed by data and statistics. 

The Upski team’s primary activities are:

• Creating an effective chatbot conversation experience

• Fostering partnerships with employers in order to offer respectable jobs to job seekers

• Managing job seekers who secure employment with the Upski program through the work processes of their employment to ensure they are upskilled to deliver top-quality products for the employers

The direct outputs of these activities are:

• Providing a wider range of job opportunities that are accessible to marginalized communities

• Executing successful job matches, which would stabilize job seekers from marginalized communities in sustainable, respectable, and well-paying jobs

In the short-term, the Upski platform has already expanded the marginalized communities’ access to jobs. 80% of our respondents have indicated our chatbot’s recommendations included job platforms previously unknown to them and that they found the resources useful. With over half of Upski’s users seeking jobs due to either COVID-19 job loss or the need for extra money, landing jobs through Upski helps this demographic secure financial stability, as well as more long-term impacts that come with financial stability, like mental health wellness. 

Additionally, since Upski works with those who are able to find jobs through our platform, we are able to offer upskilling, training, and education as they are needed. This helps our users to learn while they earn. Upskilling is critical because the McKinsey Global Institute estimated that 14% of the global workforce would have to acquire new skills by 2030 due to automation and artificial intelligence, and 87% of executives said they were experiencing skill gaps in their current workforce. Working with our users so that they continue to be competitive in the job market builds the foundation for sustainable careers. 

Finally, just by completing the conversation with our Upski chatbot, our users become part of the Upski network. Half of our Upski interviewees indicated that they found their previous jobs through a family or friend. We want to build this Upski culture of forging our community as a professional network, helping each other rise up together. 

Ultimately, these outcomes lead to our mission of generating sustainable and respectable remote livelihoods for marginalized communities.

Currently, our Upski platform has reached 1,500 job seekers. These users are now part of the Upski community -- they have received job recommendations through our chatbot and we will continuously be sharing resources and connections with them to aid in their upskilling and job search for years to come. Out of these 1,500, we have established direct relationships with 150. These 150 are Upski users that we are personally working with to secure jobs with our employer partners or other sources, or to train in order to acquire the necessary skills to land one of these lucrative jobs.

Within a year, our job recommendation system will be fully automated and more effective at pairing job seekers with tailored jobs and resources. As a result, we anticipate exponential growth, reaching 20,000 job seekers through our platform by then. This would require a growth of 1,500 per month, which we have been able to reach this past month with our pilot in Thailand. However, the one-on-one process of working with specific individuals to secure employment with our employer partners will still have a personal manual aspect. Therefore, to be conservative, we anticipate directly serving 400 of the 20,000, following the growth trajectory we are currently experiencing.

In five years, we will have a more established employer partner network that would enable us to also scale our direct relationships with our Upski community. We anticipate directly serving 20,000 out of our broader Upski network of 200,000.

Our main impact goals for the next year include:

• Increase the number of people in our Upski community who are able to land respectable jobs through our platform and connections

• Increase the number of people in our Upski community who acquire lucrative and marketable skills through our platform and connections

We plan to achieve these impact goals through the completion of the following secondary goals:

• Attain more employer partners offering well-paying respectable jobs for our Upski community

• Improve our technology to fully automate our job recommendation system so we can scale our job matching service exponentially

• Establish our methods for managing Upski constituents, ensuring top-quality performance and products

• Establish our upskilling and recruitment processes so they are more streamlined and effective

• Build strong centers in the Philippines and Thailand, where many of our users are located and we have existing partnerships we can develop

Our main impact goals within the next five years include:

• Reduce the number of unemployed/underemployed individuals throughout the world

• Reduce the number of migrant workers throughout the world who were forced to leave their motherlands just to survive

We plan to achieve these impact goals through the completion of the following secondary goals:

• Develop technology to automate even the employer partner relationship building

• Establish partnerships so that marginalized communities’ lack of access to reliable Internet connection is no longer a barrier to securing remote jobs

• Expand our reach globally by taking our model and scaling beyond the Philippines and Thailand

Main barriers to effective Upski implementation include:

• Ability to secure marketing and customer service assistance to establish and grow initial product traction

• Ability to secure technical talent to focus on product development

• Ability to secure partnerships with upskilling/education organizations and employers to collaborate on identifying new talent pools, innovative recruitment strategies, and effective distance learning

• Ability to source reliable jobs and validate ones that match with job seekers’ skills to build the foundation of a useful product and service for the marginalized communities (set stage for customer retention)

• Ability to ensure the quality of work performed by remote workers for the employers to build the foundation of a useful service for the employer partners (set stage for customer retention)

• Ability to ensure the proper infrastructure needed to truly spread beyond socio-economic borders -- establish reliable access to internet and mobile phone technology in marginalized communities 

• Ability to establish strong partnerships with the appropriate influential individuals and organizations in the Philippines and Thailand to navigate the differing cultures and laws, ensuring a strong foothold and presence in our focus countries

• Ability to take our initial models in the USA, Philippines, and Thailand, and scale globally

• Ability to secure sustainable funding sources to maintain robust technology development and operations

Our primary barriers are marketing/exposure and funding.

To overcome the financial barrier, we have set up our funding plan, which includes applying for multiple grants. We are utilizing our connections to find the best suited grant programs and opportunities. 

To overcome the marketing/exposure barrier, we are seeking talent specifically from our target countries so they’d already have an understanding of the culture and what outreach strategies would work for our target demographic. We are also seeking to partner with established organizations or influential people, as well as get major media publications from these target countries to expand our exposure. 

The next set of barriers involve establishing partnerships. 

In order to address the upskilling/education partnerships, we are using our learnings from our current chatbot, ongoing research, as well as partnership with Solve Education in order to determine what top skills would be most lucrative in the post-COVID world. This will help inform which industries are best to tap for upskilling/education organizations. We are also planning to use Solve Education’s network to jumpstart our upskilling connections.

In order to establish our employer partnerships, we are currently more focused on retention rather than acquisition. By concentrating on the few employer partners we have now, and ensuring we are able to provide topnotch service and deliverables, we can build on these genuine long-term relationships for loyal clients that could lead to referrals for growth.  

In order to best serve our target demographic and scale our implementation, building sustainable partnerships with the government, private sector, and civil society organizations is critical. Our team has strong relationships with the Obama Foundation, the MIT Media Lab, resources from Harvard and Stanford universities, and organizations that work directly with those at risk, including the Labor Protection Network. 

Our current technical and platform development partners are Codustry and Futuristic Research Cluster of Thailand (FREAK Lab) inside King Mongkut’s University of Technology Thonburi (KMUTT). These organizations provide a bulk of the technical talent and knowledge we are using to develop and deploy our chatbot product and service.

We are also partnered with Solve Education, which is an education non-profit organization. Solve Education is our upskilling partner, so if we find that some of our users need upskilling before they could land respectable jobs, we could connect them with Solve.

Our team mitigates the socio-economic damage caused by COVID-19 by using technology to help bridge the gap between businesses needing to recruit for remote positions but also needing to recover from financial losses caused by COVID-19, and job seekers from marginalized communities whose dire living situations are worsened by COVID-19. 

We have developed Upski, an intelligent job-matching system in the form of a chatbot that will be able to connect workers and employers across geographies. By connecting a global workforce in one data system utilizing the virality of the chatbot, companies and talent can be distributed globally, making it convenient for job seekers to be able to seek work out of their home countries and for employers to be able to source untapped talent globally. Our product not only recommends jobs, but also recommends upskilling opportunities, which are crucial to socio-economic upward mobility. 

Our business-facing chatbot determines open job requirements and company culture, while our people-facing chatbot assesses job seekers’ skills and interests. Upski then cross-checks the data between the two chatbots to recommend the best matches for both the employers and potential employees. With Upski, we are relieving the job seekers’ pain in the job search process, as well as the employers’ pain in recruitment. 

Our mission: steer businesses toward a more diverse workforce with a resilient structure that is more prepared for the digital future; and reach the most vulnerable of our global workforce and connect them with respectable remote jobs tailored to their needs and skills. 

In our first couple of years, our funding will come from grants. In the third year, once we are able to scale, our revenue will come from both the job seekers and employers. 

On the job seeker side, we will begin charging a small job-matching fee as a percent of the salary paid by the company that employs job seekers in the Upski network. 

We will monetize the employer side by charging employers for every job position that they would like to include in our database. With our powerful AI recommendation system, the employers will greatly benefit from using our database and service, because of the matching accuracy that we will be able to create between the job seeker and employer. The currently existing services to reach our target demographic are incredibly fragmented, plagued with job scams. On the other hand, the alternative more established platforms that may have more secure and streamlined processes do not have the ability to reach our target demographic.

Our main goal is not to monetize; our main goal is to be able to serve workers who are displaced by COVID-19 and who will continue to face challenges in earning a sustainable job and income after COVID-19. We see great potential in first focusing on the scale, followed by the monetization, with employers as our main source of revenue while we provide them with a reliable and efficient way to greatly broaden and diversify their talent pool.

We have not raised funds for Upski yet, but we are currently in the process of applying for grants to carry us through our first couple of years. By our third year, we plan to be generating revenue through our business model operations. While building our product, we have gained technical and platform development assistance partners with Codustry and Futuristic Research Cluster of Thailand (FREAK Lab) inside King Mongkut’s University of Technology Thonburi (KMUTT). 

Our plan is to raise grant money in the amount of  $300,000 by the end of 2020. This will help cover our expenses and testing necessary to grow the product, platform, and network. There is an urgency to secure this funding this year, given the magnitude of the problem as well as the talent needed to execute our critical mission to provide jobs for those most affected by COVID-19.

Expense areas 

Budget USD

Technology platform and product development 

$50,000

Marketing, PR , communications materials, and graphic design work for job seeker outreach and employer outreach 

$20,000

Human resources to support up-skilling and employer partnership 

$100,000

Total

$300,000

The financial support from the Elevate Prize will help us achieve the tech platform and product requirement for us to serve our current network of 1,000+ job seekers and expand our capability to reach even more. The challenges in job loss affect hundreds of millions globally, so with the funding, we can grow both our technical capability and our human resource team to be able to meet the challenges. We believe we have found the fit in the market and have the ability to reach the untapped underserved job seekers with our chat platform. Our vision to be able to generate a robust real-time recommendation system matching job seekers and employers will be fully developed with the funding received. 

Additionally, beyond the funding, having a platform to be part of Elevate will help us gain exposure to be able to collaborate with non-profits and other organizations and agencies that are in touch with our target demographic, as well as companies looking to find talent that could benefit from our platform.

Since Upski is in its very early stages, funding and marketing/exposure are most critical for us to gain initial traction in the market. We are all currently working on Upski on a voluntary basis, and securing funding would allow us to sustainably dedicate more time to our efforts. This would also allow us to recruit more support for our tech development and marketing/exposure -- both critical in these initial growth stages so we could build our relationship-focused network and offer top-quality service through our product. Once we establish this foundation, we could work on scaling up to expand our reach and our revenue.

Finally, as with any start-up, mentoring and/or coaching is always welcome and valuable.

Having the backing of MIT Solve and collaborating opportunities with social entrepreneurs in the MIT Solve ecosystem who work to solve similar challenges in improving economic mobility will be of large benefit to amplify the impact of our work. We are looking to connect with those who are from underserved backgrounds and are in need of upskilling and securing a sustainable income source. We would look to partner with the Ministry of Labors of different countries as well as the corporate sector -- SMEs and Fortune 500 companies -- who are currently hiring. We are also keen to establish partnerships with companies that have a strong social mission to upskill workers, like Microsoft through LinkedIn Learning platform and Coursera.