Rani Jobs (by Myna Mahila)

Today's job marketplace does not harness the full potential of women and girls at the bottom of the pyramid. While existing platforms connect employees to employers or tasks, there is a mismatch of literacy skills, aspirations, mobility requirements, and technology access for low-income women, leaving them to work service-based, laborious, low dignity jobs. Rani's micro-tasking platform provides jobs to women in a flexible, fun, and dignified way. We source job tasks directly from private sector clients in artificial intelligence and machine learning (dataset generation, data labeling), social media companies (content creation, categorization, and search optimization) and research labs. Clients upload their data into a back-end API and specify parameters for accuracy, turnaround time, and redundancy. Rani assigns price, breaks the data into micro-tasks and pushes them onto the smartphone application. Women can also upskill in sector-specific and language skills, and connect with each other in this worker-centric job tasking platform. 

With merely 23% of women in the labor force today in India, despite advances in education levels, it remains puzzling why high willingness to work does not translate to employment. Women lack job opportunities that meet their needs: part-time, flexible, at or near home, and safe and dignified work that their family would accept. Today's job marketplace in India does not harness the full potential of women at the bottom of the pyramid. While existing platforms connect employees to employers or tasks, there is a mismatch of literacy skills, aspirations, mobility requirements, and technology access for low-income women, leaving them to work service-based, laborious, low dignity jobs. 

Especially with the COVID and post-COVID job scenario, the urban poor, and particularly the vulnerable women are most likely to be left out. Women are often encouraged to simply do jobs in handicrafts related work, which are mostly low paying. Further, more educated girls prefer not doing any job at all to doing these lower paying jobs. In fact, labor force participation rates (LFPR) are the lowest among women with secondary and higher secondary education, while they are the much higher among the less educated (~53%) and college educated (~28%). However, women who are "middle-skilled" are least likely to find a suitable job. 

As education levels rise, but only till secondary schooling, the prospects for FLFP look bleaker in India. There is a desperate need for jobs that support these middle-skilled women so they can enter and stay in the labor force. 

Let us take Nisha’s case. She is a young, high school educated girl in a slum looking for a job, but she has to take care of her children at home. She is among the 100M millennials in India who want a flexible job at home.

In our interviews with more than 2,000 women in Mumbai slums, we have heard repeatedly that a platform for women must solve for four issues - flexibility, community, relevance, and safety. 

• On flexibility, women want part-time, work from home jobs because of social mobility constraints, norms, and childrearing responsibilities. As a smartphone-based app, Rani enables this.

• On community, women want to feel supported by a network of peers and mentors. Rani provides an in-app integration with Whatsapp to enable women to build this network and ask questions to each other as they complete tasks. 

• On relevance, women don’t want to just earn; they also want to learn skills that might help them achieve more respect in their communities, such as English. We are ensuring that our tasks are coupled with a robust upskilling curriculum to build these skills and improve performance.

• And lastly, on safety, women don’t want to worry about harassment on platforms from men. We are resolving this issue by limiting the work to home or in local centers next to their homes, and heavily investing in cyber security.

As a women only platform, Rani uniquely leverages our understanding of women’s needs to recruit and retain a large workforce of semi-skilled workers.

Rani addresses semi-skilled women's lack of employment opportunities in India with the digital jobs of tomorrow. Our mission aligns closely with that of the Challenge to identify and support workforce development solutions to provide employment opportunities and upskilling for some of most underserved populations in India. Our focus is on low income women, living in urban slum communities, peri-urban or rural areas where smartphone penetration is now rapidly growing. 

As the book "Ghost Work" outlines, an increase in automation leads to increases in human verification work. Rani Jobs connects these increasing digital jobs by companies that focus more on automation and machine learning, which need huge amounts of data captured and then verified, with the workers who need these types of work the most. While workers need to be upskilled in learning more digital tools, upskilling without clear employment opportunities may not be as effective, as numerous studies have shown (particularly in India where after completing even a year long skilling course in IT, more than 80% of the youth are unemployed). Being worker-centeric, and aligning with all three goals of the Challenge, Rani Jobs increases female labor force participation in India among low income communities in doing digital jobs, which are immensely scalable and sustainable. Further, it creates a community of workers connecting their skills and upskilling them further to take on more complex work and build their careers in ways unimaginable before. Finally, it connects skills of this workforce with employer needs for different types of work available. 

Suhani Jalota, Founder at Rani, Founder at Myna Mahila Foundation

Human workers doing back-end work that ultimately feeds into machine learning algorithms is an established space of work. There are thousands of companies to whom this type of work is outsourced, and yet there is more work to be completed everyday than there is capacity to complete. Tapping into the same workforce is likely not sustainable or scalable. This is where Rani steps in. Rani's innovation lies in designing a platform that uniquely onboards an entirely new group of workers - women with household constraints and huge financial needs - to meet the market demand for workers to do such work. This is done by converting the traditional office model to work-from-home and in newly established Rani women-only centers in their local communities (5 minutes walking from home), and shifting PC-only work to tasks over smartphones to be widely accessible. 

Our second important innovation lies in creatively converting large companies to provide their lower skill tasks to new workers rather than pay hefty fees for people who are not required for this work. For instance, insurance companies could easily get images of insurance claims and marketing agencies could cut on costs and get photos clicked of banners and adverts by the currently unemployed youth and women in local slum communities instead of by their salesmen and highly paid marketing managers. With millions of women newly coming online and having access to smartphones, there are many unique capabilities they bring that haven't been tapped into before. 

We use technology at three levels. (1) Our worker-facing product is a smartphone based application that provides jobs on smartphones to women and girls in a flexible, fun manner. How does this work? We source jobs directly from private sector clients in artificial intelligence and machine learning (dataset generation and data labeling) and social media companies (content creation, categorization, and search optimization). (2) Clients upload their data into a back-end API and specify parameters for accuracy, turnaround time, and redundancy onto a client-facing portal. This portal includes advanced analytics about the accuracy and timings of the work completed along with the worker profiles. Clients know in real-time about the progress of their outsourced tasks, and can connect with our Supervisors/Managers regularly. (3) Using our Administrative control, Rani assigns the price, breaks the data into micro-tasks and pushes them onto the application. How do we make this work fun? Tasks on the Rani app are gamified (e.g. multiple rounds of progressive difficulty, three-star achievement system, leaderboards, in-app coins to world build) to make repetitive tasks fun, which can improve retention and accuracy. How do we ensure accuracy? Rani uses built-in quality assurance algorithms to ensure accuracy and relies on automated task redundancy (e.g., each task completed by 3-4 users), which can boost individual accuracy of ~84% up to platform accuracy of 99.84%. In the future, we plan to build some more automated checks on the tasks, making the process even smoother. 

Rani's theory of change is as follows: With upskilling courses on the platform and the direct employment tasks as the inputs, women can now become more digitally literate and join the labor force. Since the jobs are now provided in ways that are suitable for women at home or next to their homes, they are more likely to accept employment along with their family's consent. In a randomized control trial (RCT) we are conducting with Stanford University with 600 women in Mumbai's slums, we find more than 60% of women who were previously unemployed are now accepting employment through Rani due to its convenience and dignified work even at very low payments. This number is at 80% when payments are higher. Once they start to work (~20%), ~90% of women continue to consistently work on the platform. They then spread the message to others in their community to join.

The outcomes are an increase in household incomes, confidence, agency and dignity. The RCT shows how women are now earning 4x that of what they could have earned elsewhere. Women claim to "never have felt this way before" and they can now "show my husband and in-laws what I can really do"! Within a week of working on Rani, women were buying smartphones and proudly telling others in their native village even about their new job. Ultimately, we want the impact to be to shift social norms for women and decrease their stress levels to help them realize their full potential. 

Over the next year, our goal is to provide employment to 3,000 women (and complete the RCT (phase 2)) and observe the effects of Rani on their agency, confidence, social norms, dignity, mental health and overall livelihood indicators. We would contract with at least two large clients for providing ~300,000 hours worth of work on the platform for the larger pilot. We expect to see dramatic improvements in women's confidence and agency scores. 

Over the next 2-3 years, we will work with partner NGOs, corporates, and other community based organizations who are looking for employment opportunities for their workers and their families. We already have an estimated 120,000 women in four states in India (Gujarat, Bihar, Karnataka, and Maharashtra) waiting through such collaborations, but we cannot proceed until we have sufficient and consistent work for them on the platform. We will further develop the upskilling courses and be able to measure our impact based on increases in digital literacy and performance on the platform. 

In the next five years, we will try to collaborate with local governments, for instance with the National Urban Livelihoods Mission (NULM) where there is currently a huge dearth of employment opportunities for the tens of thousands of women currently unemployed but looking for work. The government partnership could help in scaling up women-only centers in urban slums and villages. Our goal is to have a Rani women's digital jobs center located in every slum community and village in India, providing sustained employment at scale. 

Rani Jobs is promoting UN SDG-8 of full and productive employment and decent work for all and SDG-5 of gender equality by creating a job platform for women who are left out of the economy. Hence, in the short run, we measure the number of women who have newly joined the labor force, those retained, and their incomes. Further, we look at the benefits of employment on their own life and for their families. This includes a PHQ-9 metric for measuring mental health changes, a tension scale, women's agency metric, and an adapted dignity tool from healthcare. Confidence and aspiration levels would also be measured through commonly used tests that calculate scores and can see changes over time. 

In the longer term, we may expect to see changes in social norms (measured by agreement to statements such as "My husband thinks I can manage my household and work in a job.") as a result of seeing women work for the first time. 

We are very serious about collecting rigorous impact data at Rani. We have a long term RCT running, and are capturing data from both wives and husbands about women's work on the platform at continuous intervals. 

Meeting next year's goal of providing employment to 3,000 women is dependent on having enough client work available to provide them with stable income. Clients are used to providing work over PCs, so they are hesitant to provide smartphone-based work. Many companies are unwilling to outsource work to our Rani software over smartphones, and some want workers to come to their office instead of doing the work from home or nearby local centers due to data privacy concerns. This traditional view seems to be hard to break. 

Rani Jobs is a two-sided marketplace, so workers stay so as long as there is work on the platform, and work comes if high quality workers are using it. While there is excess labor supply, we are constrained by labor demand. If workers don't perform up to client's expectations, they may not continue contracting with us. This could leave many workers without jobs for some time. 

Labor laws in India are not as favorable for employers; With labor policies for gig workers not yet fully established, there is a risk that they may become more stringent to become too expensive to keep a "dormant worker base" when less work is available. 

We are doing many client-side interviews, and are looking for dedicated Advisors from the Business Process Outsourcing industry in India, and from ML/AI teams at large technology firms. We are exploring different methods of contracting - reaching out to BPOs to contract to us, and directly to private firms who need their tasks completed. Further, we are consulting with software experts to ensure our system is fully protected and data privacy is not much of a concern. 

We need to invest a lot of time and resources on creating the right incentive structure for paying workers that is in line with company success and client satisfaction. This includes understanding how, if any, should a fixed payment structure work, how it can be combined with incentives and/or penalties and what the payment is pegged to. If clients are most concerned by accuracy, we would need to pay based on high accuracy levels, and not by total tasks completed or time spent on the platform. We will consistently increase our bar for accuracy (till the threshold). 

From the beginning, we are studying labor laws carefully, and will hire consultants to support us in creating a robust system for Rani workers. 

Rani Jobs is currently an initiative of Myna Mahila Foundation, which is a registered Section 8 or nonprofit company in India. Myna Mahila is housing this initiative from incubating the software development to providing the team for recruiting and managing Rani workers on the platform. 

Apart from 60-person Myna team, 15 full-time members: Myself as Founder/ED, Program Manager, Data Associate, two Quality Assurance Supervisors, Accounts and Payments Executive, and nine field coordinators to support the Rani workers. 

Software development is contracted out to an agency where we work with two full-stack engineers and UI/UX designers. 

Our team is a good balance of those who resemble our worker profile, and those who come from the data industry and understand the type of work we contract. Most of the team is women (all except one), and many come from low income communities from slums where they are the first women to be employed from their families. They understand the value of earning income and being independent, and they carry forward this commitment to work. 

We have temporary centers within different slum communities where we interact with workers on Rani daily. Further, my team and I go door-to-door to visit each worker, from the star performers to those who never open the application, and try to understand why they love or dislike Rani. We are in constant touch with them over Whatsapp, calling, and home and center visits. We also hold focus group discussions with groups of women to discuss specific changes or additions that need to be made to the platform. Everything is ultimately decided based on what the workers need. 

We have excellent team members working closely with clients, trying to understand every challenge that could come up and build for it in our design. 

It is crucial that our team members complement each other well and bring a variety of skillsets and perspectives to the table. We ensure diversity from top down in our cultural and economic differences. Further, we encourage perspectives to be heard in bottom up order. During team debriefs, people working most closely with workers speak first and they are given the most air time. We constantly relate back to our values during the meetings every time there seem to be diversions, and we regularly engage in open and honest dialogue (including mistakes/failure). If our team believes in diversity and inclusivity themselves, we would be equitable in recruiting workers of different ethnicities and cultures for Rani. We provide training to be unbiased in the recruitment process to ensure we are not discriminating. For instance, our Muslim coordinators visit Hindu homes to speak about the job, and vice versa. When our diverse team talks to each other so openly and freely about their lives, others around them tend to open up as well. 

Finally, we need to invest a lot more in leadership development of our team members, providing them with mentorship to grow more confident of themselves and their team. 

We work with the rest of the Myna Mahila team and PUKAR (another local NGO that supports hundreds of thousands of families in slums) to build trust among beneficiaries where they have worked. We also partnered with Keerti classes, a computer training institute, for digital skilling courses. Further, along with Software Technology Parks of India (STEPI) and the Ministry of Electronics and IT (MeritY), Govt. of India, we are reaching BPO clients in India. 

We are also working with Stanford University to conduct the RCT around Rani to see how effective it is at providing women with sustained employment.  

Rani operates in the intersection of multiple stakeholders: (1) Workers need income and new skills and don't have "suitable" jobs available that allow them to earn income flexibly and reliably, (2) Employers (BPOs or direct private technology company clients) need high-quality reliable workers that can be retained, and need un-biased algorithms since existing statistical fixes for debiasing algorithms won’t be able to make up for the lack of labeler diversity, (3) NGOs/Corporates need to provide their workers and their families with monetary support, (4) Local Governments need to somehow create jobs for tackling the massive problem of joblessness among women, (5) IT Skilling agencies have poor rates of employment post their skilling courses and need to better guarantee results. 

Rani is able to meet the needs of all the stakeholders, particularly those of the workers and employers, creating a win-win situation for everyone. 

Most of our current challenges are client-facing: Which clients would be willing to take a risk with us and try out a smartphone-first application with women from home? How can we make our product more fitting for clients' needs? Does the solution need to be customized for every client or should the go-to-market strategy be a blanket solution for now? 

We would seek access to mentors and coaches who know the Business Process Outsourcing and machine learning/AI data input and labeling industries well, and can help guide us to anchor on an effective model for clients, along with the existing worker-centric solution. 

Further, maintaining workers' high quality standards and incentivizing them effectively are also challenging for the platform's sustainability. It would be incredibly useful to explore various models with mentors for building an incentive structure that matches clients' quality expectations perfectly. We would also benefit, as a team, from the peer-to-peer network and the impact measurement practice that can ensure we are rigorously monitoring our metrics. 

Funding from this challenge would be also instrumental in helping us develop our technology solution and our backend verification algorithms further, which is another bottleneck for the moment. Funding constraints have stalled further software development (as we continue to pilot the existing version with workers), so funding from this Challenge could help us get going more quickly on making major updates to our existing applications and build other customizable software. 

Currently, the client-facing or service distribution area is the number one priority, as we are looking for partners who can provide us with work on the Rani platform. It would be immensely helpful to have some client leads that could be interested in working with us. 

A close second and going hand-in-hand with the service distribution area is technology/technical support as the software updates we have to make to RANI's platform may depend heavily on the type of work we get on the platform. Here, we would like support in guiding on the best path to build a seamless software solution that combines clients, Rani admin/verifiers, and workers on one platform (or multiple). 

Third is human capital, for hiring technical team members who have worked closely with AI/ML training datasets in the past. Any suggestions for how to go about hiring technical talent (and whether this should be in India) could be of great help as well.