Jijenge Academy

We equip vulnerable youth with the tools to break the cycle of poverty by teaching them a monetizable, digital skill to achieve financial security. With a 90% job placement rate and 4x average income increase, our program trains young women with no computer literacy to become advanced data labellers in two months.

We corner an untouched market gap in Africa: Enabling base-of-the-pyramid distributed workforces. We do this by 1) building cyber workstations to facilitate cheap access to high-specification laptops and high-speed internet, and 2) equipping third parties to deliver our training through a social franchise model. 

If scaled globally, our infrastructure can permanently replace dangerous manual labor work with data annotation work for the poorest women in the world. Doing that would catapult women into the digital economy, formal sector employment, and start them on the path of bridging the huge gaps created by education inequality in emerging markets. 

Kenya has 10 million youth, 65% are unemployed. Most underserved youth earn through informal employment of <$50/mo in dangerous casual labor jobs such as tea plantations. Women face more challenges than men due to sexual harassment and lack of sexual education. 

Young women have no tools to achieve financial security, education, or dignified work. This problem is compounded by lack of access to finance, and the lure of prostitution or early marriage to provide for basic needs like food. 

We believe this magnitude of unemployment and lack of education can only be solved by leapfrogging traditional education and career paths, and directly training the most underserved women in a monetizable, digital skillset. That brings us to data labeling, which we believe is the most time and cost-effective digital skillset for women to achieve financial security.

Data labeling is the main bottleneck to the growth of machine learning, and its market is expected to triple to $5 billion by 2023. Kenyan data labeling companies have emerged as a result, however they haven't been able to run distributed workforces. Due to infrastructure challenges in East Africa, they have to run in-person managed workforces that commute to their HQ.

Our solution features three components that work together to help vulnerable girls break the cycle of poverty they've been born into.

1) Our training program takes young women from no computer literacy to advanced data labeling skills in 2 months. After 3 years of iteration, it is tailor made for those from the poorest backgrounds to succeed. Our community engagement strategy to source the most vulnerable girls, anticipate their needs, and help them succeed is a key part of our product offering. The curriculum also includes general employability skills, teamwork skills, and sexual education and health. 

2) Our low income cyber workstation infrastructure is desperately needed to enable distributed workers in Africa. We charge workers a weekly access fee, which allows these workstations to pay themselves back after 12 months. After that, they become remain profitable for their remaining 2.5 year lifespan. 

3) Our international data labeling employment partners supply online data annotation work. Thus, women can begin earning directly after the training program. Our B2B product is building affordable, distributed data labeling workforces for international companies and allowing them to take it in-house or keep paying us to manage it. 

We formed this venture (and our team of co-founders) by listening to the needs of high school graduated orphans from the slums. We used their feedback to form a sustainable solution for them to break the cycle of poverty in a lasting way. We came to a niche that was such a value add to their lives, that our dropout rate is non-existent.

By teaching computer literacy and facilitating job access, women quadruple their income and access online resources to better their lives. After the program, students use their income to attend school part-time, move their families out of the slums, or sponsor their siblings to attend school. This creates a permanent and sustainable shift for vulnerable youth to help themselves, without external support. The best evidence of the life-changing impact we've achieved is in the autobiographies of our graduates (found on our website).

The success of our workstations and trainings depend on a deep understanding of the challenges vulnerable youth and workers face, and mobilizing surrounding communities to enable their success. Our operations managers dual-function as community champions who familiarize themselves with community leaders and the most ambitious, needy youth of communities in geographic proximity to our workstations.

Although in the past our program has had a 50-50 split of young men and women, we recently had our first girls-only cohort of women from the second largest Kenyan slum. We'd like to continue focusing on women if we win this challenge.

Our curriculum significantly strengthens girls' digital literacy, and prepares them for formal employment (dimension 2).

We board students so they can study in a safe environment, far from the distractions and dangers of the slums (dimension 3).

If scaled, our program would increase the number of young women that are participating in formal trainings. (Dimension 1)

We are the first to do what was previously unworkable: Quadruple the income of the world's poorest, most vulnerable young women in 2 months. Currently, there is also no solution that 1) enables distributed data labeling workforces in Africa or 2) offers affordable high-specification laptop access and high-speed wifi access to low income workers.  

Our closest competitors are Samasource and Digital Divide Data. They manage in-person workers who commute and solicit for data labeling contracts to staff their employees. Both of these companies must accept workers that already have a strong foundation in computer literacy (which generates less impact) and run costly in-person managed workforces since 1) their work streams are not compatible with distributed work and 2) they have no affordable solution to facilitate worker's access to high speed wifi or high specification laptops. When they look upmarket for workers that already have these requirements, there is a salary mismatch with middle class workers which leads to high employee turnover and low motivation.

What is innovative about our revenue model is that we are not competing with data labeling shops, we build the infrastructure to enable them to run distributed workforces and reduce their costs. Since we derive our revenues from fees the workers pay to access our infrastructure, we have a cost-affordable edge over every other consultant or BPO workforce builder because we can undercut their pricing because of our secondary revenue stream.

The most important technology is the data labeling platforms workers earn on. Currently, the majority of our data labeling work comes from the remotasks platform, from a workstream called 3D Lidar Annotation. Lidar stands for light detection and ranging, and the 3D data labeling our workers do in this area help automate self-driving vehicles for companies such as Tesla and Google. We also train in image annotation, data transcription, and many other work streams. 

Second, our technology-enabled cyber workstations feature laptops and high speed internet which is required for workers to accomplish a high volume of tasks while maintaining a task accuracy that allows them to earn the maximum salary. Technology also allows us to monitor the hours, quality, and tasking speed of our workers, which our supervisors are responsible for among other things like keeping internet companies accountable for delivering the paid-for bandwidth.

Our training program in computer literacy, general employability skills, and data annotation also relies on an LMS platform to distribute. We are in the process of digitizing our 2 month training curriculum so that we can minimize teacher involvement in the basic computer literacy segment. After optimizing the training into a low-touch model, we equip schools and computer labs across the country through our social franchise training toolkits. This allows them to run the training program and provide job access to their community members.

We have trained 68 vulnerable youth over 2 years (39 women), proving that the poorest young women with only a high school degree can do advanced data labeling work. We've placed these workers at a 90% job rate, with a 4x income increase. Within 6 months of job placement, 30% of our students are promoted to leadership or reviewer positions. Testimonies from our employment partners will demonstrate that our graduates have higher than average productivity, and very ambitious work ethics compared to college graduated workers.

We have recently set up our first 17-seat cyber workstation of 34 workers in Nairobi, Kenya. The cyber workstation is running day and night shifts of workers that have been earning an average of $40 per week on the remotasks platform in 3D lidar for the past month. The validation proof that we are an effective workforce builder can be seen in our partnership with Scale AI (the parent company of remotasks). This Silicon Valley startup successfully trained thousands of low income, distributed workers in the Philippines to earn via their distributed work platform. However after almost a year of training thousands of Kenyan workers in 2020, they had only reached 20% of their desired work output. When we piloted training 17 of our workers in remotasks lidar annotation, Scale AI was amazed that we were producing 13% of the country-wide output on their platform. Our success is due to our deep familiarity of the East African market and the challenges low skilled workers face.

Our venture’s activities can be grouped into two areas. The first is a 2-month training program which includes computer literacy training, general employability training, sex education, and advanced data labeling skills acquisition. 

The output of the computer literacy program includes access to beneficial knowledge and online programs. By teaching girls advanced web browsing, we ensure they understand they can access unlimited information and e-learning portals to upskill themselves. An output created by the general employability training is career progression since we teach how to find jobs online, how to apply, and how to professionally communicate. Proof of this includes our graduates who have moved into higher paying jobs or shortlisted for jobs with salaries 3x theirs. An output from our data labeling skills training is that women develop a monetizable skillset alongside other job skills such as attention to detail and problem solving. An output of sex education is an increased understanding of concepts like consent, agency, and birth control. 

The outcomes of training vulnerable young women in data labeling is a higher digital literacy among women. The outcomes of access to beneficial knowledge and online programs are better health practices and better self-education. Outcomes of understanding agency, consent and birth control is lower rates of teen pregnancy and early marriage, and better sexual health. 

Our activities connect young women with data labeling jobs. They begin tasking on real work near the end of the training and are earning by the time they graduate. This creates two outputs: 1) formal employment with promotion opportunities down the line and 2) girls graduate our program with a 4x income earnings increase.

The outcome of increased income is financial security for the poorest women. After quadrupling their earnings, they can afford their own housing outside the slums, basic needs, and are more resistant to financial shocks so they can save to achieve their goals. We believe lack of financial security is the biggest contributor to women not reaching key learning milestones due to the pressure to provide for basic needs, which involves getting married early or pressured into prostitution. 

We have trained 68 students to date (39 women), achieving an average 4x earnings increase upon graduation. In one year, we aim to create employment for 1,000 women by operating 16 cyber workstations across East Africa. In 5 years time, we expect to have created almost 500,000 jobs for women across Africa. 

The majority of that number will have been achieved via our social franchising model by distributing data labeling training toolkits to computer labs across the continent to onboard workers onto self-regulated data labeling freelance platforms like Scale AI who automate pay, quality control, and new skills training. Although our estimates are conservative, we believe we can upskill millions of women in distributed work by building an LMS tailored for vulnerable women in rural and urban areas. 

The rest will be achieved by soliciting direct contract work to build affordable distributed workforces for international companies, or charging data labeling companies to use our infrastructure to run cheap distributed workforces. 

Our goals within the next year are to raise a seed round to train 1,000 workers across Africa (we have secured the work supply already) and do the following: 

• Invest in marketing and sales to solicit contracts from organisations interested in building in-house or distributed workforces

• Pilot blended learning to fully digitize our social franchising training toolkit and begin distributing it to partners

In the next five years, we will have achieved the following goals:

• Successfully built high quality, African data labeling workforces out of the most vulnerable youth for large, international clients in East, West, and South Africa (creating ~75,000 jobs), and looking to expand to other geographic regions outside of Africa

• Spread our social franchise toolkits & LMS platform to ~425,000 women by marketing to schools and computer labs and equipping them to train community members 

The cyber workstation segment is how we achieve strong cash flows, and requires mastering operational complexity. However to reach millions and achieve scale, we must partner with other organisations via a social franchising model.

There are so many initiatives that promote technology in Africa that are being under-utilized because they are simply teaching  internet connectivity for knowledge acquisition instead of breaking the cycle of poverty. Given the abundance of data labeling platforms that automate quality control, pay, and job-specific trainings, it is possible for us to equip millions to access this work after they go through our training, which is is uniquely designed for youth from the poorest backgrounds to understand.

The barriers to us reaching our goals in the next year include if:

• We are unable to raise at least $350,000 seed round to accomplish our goals next year

• Covid-19 materially affects Kenya in a way that requires all businesses obeying social distancing rules to temporarily shut down for an extended period of time

• There is a material change in the supply of data labeling work available because large suppliers such as Tesla or Google declare bankruptcy

The barriers to reaching our goals over the next 5 years includes the following:

• Every African country we expand to has cultural differences that we must fully understand in order to build trust with communities, help the most vulnerable youth succeed, and tailor our pedagogy in a way that gets through to them

• Under-estimating or under-sizing the global demand for distributed data labeling workforce building would require us to pivot that segment of our business model to find utilisation of our cyber workstation infrastructure in another way, which would take extra funds and time

• A challenge of African expansion is that suppliers, employees, and beneficiaries in this market consistently try to take advantage of organisations perceived to be funded or connected. This is amplified in informal or low income areas in Africa where there are less formal procedures to get business done, and it can be a major obstacle in achieving optimal unit economics.

For barriers in the next year:

If we are unable to raise a large enough seed round, we will scale back the number of cyber workstations we plan to build, and continue as we have been to reach or goals on a shoestring budget at a slower rate.

We researched the business social distancing regulations and upgraded our workstations to be covid safe. We also engaged authorities to confirm we are abiding laws if covid escalates. Since we generate good salaries for 64 members of each community we operate in, we would be the last business asked to shut down in dire circumstances. 

If our partners' data labeling clients declare bankruptcy, we will seek out new data labeling work suppliers. We have an edge because if needed, we don’t have to charge a premium to companies since we generate revenue from the utilization of our workstation.

For barriers in the next 5 years:

In our expansion countries, we must always acquire talent that has a deep understanding of cultural differences, especially related to the challenges workers face and their community contexts. Without this, it would be impossible to tailor our pedagogy.

If we find that there is not a high demand for building cheap distributed workforces, we will have to explore an adjacent niche for our distributed infrastructure.

In response to price sabotage in low income areas, we mitigate this by hiring trusted & well referenced team members in new geographies use checks and balances to maintain an honest culture.

Full time staff currently include 4 employees: Operations Manager, Head Trainer, and 2 graduates who supervise the night and day shift of our cyber workstation.

Un-paid part time staff include 2 board advisors, 2 volunteers, and our partnership development manager.

The CEO / founder is currently a full time contractor, but over the past 2.5 years was a part-time, unpaid employee.

Our team is the ideal combination of business acumen and expertise working with vulnerable communities. I am an ex-JPM Investment Banker who moved from NYC to Nairobi with an East African investment advisor to consult and raise money for East African SME's and startups. Afterwards, I moved to an ed-tech startup as Growth Manager to help them refine their business model for Pan-African expansion. My skills include development of sustainable business models, intercultural communication, and finance.

Our chief board director is named Jacob Reisch, a Forbes 30-under-30 successful CEO of a series C Boston tech startup whose skills include post-Series C growth, fundraising, and management.

Our operations manager is a community champion for the underprivileged. He was the healthcare provider of an orphanage where he implemented holistic plans to build clinics, zeroed mortality rates, rallied foreign service officers, and fundraised for the orphanage to triple its size.

Our head teacher and trainer used to work at Samasource, one of the data labeling companies we placed our graduates at. He leads our training program and curriculum development. His skills include data labeling, training underprivileged youth, and quality control. Since he came from an underprivileged background himself, he is best positioned to take students from no computer literacy to high quality data labelers.

Our partnership development manager, Jayne Shah, worked as a project manager at Cloudfactory, one of the largest Kenyan data labeling companies. Her skills including managing the client-BPO relationship, quality control, and sourcing data labeling employment partners.  

We are currently partnered with Scale AI, a silicon valley startup with a large, untapped supply of 3D lidar annotation work. Scale AI’s platform for distributed workers in emerging markets is called remotasks. We train the most underserved youth in lidar annotation for remotasks, and facilitate their continued access to data labeling work. We work with remotasks in the following areas:

·  Promote high quality taskers into reviewers, and subsequently super-reviewers

·  Train our staff to be experts in 3D lidar annotation

·  Data insights & monitoring into our workforce’s hours, tasks, accuracy, and pay

·  Piloting the effectiveness of varying pay schemes such as hourly, performance-based, etc.

Our business model provides value to 2 types of customer:

1) We train vulnerable women in underserved communities to acquire a technology-related job, and skills to uplift them out of poverty

2) We build affordable, distributed workforces for large, international companies that need data labeling work, or data labeling companies that need distributed workers

We train and facilitate permanent access to data labeling work for women in one community at a time by renting a space and furnishing it with a minimum of 32 seats & laptops, as well as high speed wifi. After 12-14 months, each cyber workstation pays itself off and generates profit for the remaining 2.5 years of its life. The lifetime revenue per worker is $1,070 kes, while lifetime costs per worker are only $620.

We begin with a 2 month training program in the workspace, that trains all 64 workers (if it is a 32 seat venue) in computer literacy, general employability skills, sex education, and data labeling skills. Workers begin earning directly after the training in a day shift and nights shift. Workers are paid each week, and pay us a weekly access fee to use our equipment. We partner with a data labeling employers to tailor the data labeling training to their needs.

Attracting the most ambitious but underprivileged workers in a community is one of our core competencies. This is why community engagement and geographic proximity of our workstations with communities is a core part of our model. 

Our financial model projects breakeven in year 3, and revolves around operating cyber workstations in underserved communities that train and connect the community with data labeling companies in need of affordable, in-house workforces. Our revenue comes from three sources:

1) Each worker that utilizes the cyber workstation pays a $7 per week access fee 

2) Companies looking to build in-house data labeling workforces pay Jijenge to build them an affordable distributed workforce, or data labeling companies pay to access our infrastructure and run distributed workforces

3) Social franchising our digitized training course and data labeling toolkits to schools and computer labs across Africa. This generates revenue via franchise fees to access the toolkits to onboard workers.

We plan to raise investment capital to cover expenses in year 1 and 2 before we can win larger contracts and achieve breakeven.

The most exciting thing about getting selected as a Solver is tapping into MIT's network of tech alumni. The data labeling space is a competitive space to break into, with lots of international players. It has been difficult to get informational interviews with data labeling companies to try to size the demand for data labeling workforce building. One of the last validation points of our business is if catering to that niche will produce the revenues required to become profitable, so getting help to figure that out as early as possible would save time, money, and resources. 

The prize funding is exactly what we need to catapult our expansion and replicate our success. After growing enough to produce $50,000 in revenue, we could open the gateway to raise equity funding and complete our pivot to a highly scalable, high impact growth startup. 

The marketing exposure and ethos from being as an MIT solver would also help us form connections, fundraise, and acquire large contracts to train workforces.

We would love to partner with companies or experts in the following areas:

• Distributed data labeling platform such as remotasks who are looking to expand in countries with cheap labor pools

• Grant writers or experts to maximize the amount of grant capital we could achieve to start pilots in more difficult African markets 

• Graphic designers, SEO optimisers or developers to help optimize our online presence to solicit data labeling contracts

• LMS platforms for us to partner with and piggyback off their platform instead of building our own to distribute our social franchising training toolkits

• Impact measurement consultants to quantify and expand analysis on the anecdotal impact evidence we have

• Experts well connected in the data labeling industry to offer wisdom and make us aware of the micro-niches and how much competition there is in each

We would like to provide services to organizations that need cheap but high quality data labeling workforces. Thus we would take advantage of MIT's network of alumni to source companies that would want to pilot this service with us. 

Additionally, we would want to partner with MIT faculty or initiatives to test new applications of machine learning used in experimental technologies. In this partnership, we could use our impact workforce to train a high volume dataset in a short period of time to assist the researcher in honing in on the accuracy and quality of the AI. 

We would also like to find a mentor among MIT faculty who are familiar with the international data labeling space, so we can build off their learnings in order to size the market and understand the micro-niches within that market. Since the sector is highly competitive, it is quite difficult to conduct market research without connections.

Jijenge recently had its first female-only cohort of students. We board these girls so that they can study in a safe environment, far from the distractions and dangers of the slums. The core mission of our program is to help vulnerable youth break the cycle of poverty by achieving independence and financial security. One of the ways we have seen this happen for the women in our program is the following:

1) By rescuing them from an environment where they are being sexually harassed or pressured on a daily basis to get married, they are able to focus on skills acquisition and self-development

2) By the end of our program, when they are earning from well-paid work, they are able to permanently support themselves to live outside the slums. This gives them much more control over their lives, especially regarding when they will get married since the pressure of providing for basic needs isn't there.

3) Our curriculum includes sexual education and health, where our female workers learn for the first time about concepts like Consent, birth control, the signs of a healthy vs unhealthy relationship, etc. In this training, we have uncovered various family situations such as incest rape that require intervention from our community champions. Thus this education also improves the lives of workers' siblings who aren't in our program.

If we win this prize, we will use the proceeds to continue to have female only cohorts, instead of returning to our 50-50 split of men and women.

Jijenge recently had its first female-only cohort of students. We board these girls so that they can study in a safe environment, far from the distractions and dangers of the slums. The core mission of our program is to help vulnerable youth break the cycle of poverty by achieving independence and financial security. One of the ways we have seen this happen for the women in our program is the following:

1) By rescuing them from an environment where they are being sexually harassed or pressured on a daily basis to get married, they are able to focus on skills acquisition and self-development

2) By the end of our program, when they are earning from well-paid work, they are able to permanently support themselves to live outside the slums. This gives them much more control over their lives, especially regarding when they will get married since the pressure of providing for basic needs isn't there.

3) Our curriculum includes sexual education and health, where our female workers learn for the first time about concepts like Consent, birth control, the signs of a healthy vs unhealthy relationship, etc. In this training, we have uncovered various family situations such as incest rape that require intervention from our community champions. Thus this education also improves the lives of workers' siblings who aren't in our program.

If we win this prize, we will use the proceeds to continue to have female only cohorts, instead of returning to our 50-50 split of men and women.