TaQadam

Economic inequality, conflicts, and mass displacement resulted in the unprecedented youth unemployment globally, with 60 million jobs to be created to stabilize the situation in the Middle East alone. Much of the public sector and UN funding is directed towards ICT skills training for youth in Lebanon, Iraq, and Jordan, yet conflict-affected economies cannot absorb and utilize the produced talent leaving thousands of educated youths as subjects of exploitative informal economies. 

But how a young refugee with a smartphone is different from a digital nomad? TaQadam engages youth in post-conflict communities in optimizing image annotation for data-driven companies with visual AI. With an API and a cloud architecture, we ensure a simple and secure way to build image data set with a high accuracy and precision.

Mass displacement resulted in the unprecedented youth unemployment, with 60 million jobs to be created to stabilize the situation in the Middle East alone. Associated economic and social costs are immense for host communities: https://bit.ly/2pvfRd9. Despite numerous training programs in digital skills in Lebanon, Jordan, Iraq conflict-affected economies cannot absorb and utilize the produced talent leaving thousands of tech-savvy, educated youths unemployed.

We apply our AI-innovation in the setting of geospatial innovation as a focus area. Accurate geo-coded asset data enhances the ability of companies, regulators, governments, the UN and civil society to manage programs, measure risks and vulnerabilities, design marketing strategies, and optimize supply chains. By securing data at higher refresh rate, our datasets allow users to develop alternative proxies or analytics. As a company’s vision, in TaQadam, we believe the development of a global catalogue for every physical asset in the world is technically feasible. Identifying assets (e.g. mines, farm containers, schools) and assets’ features (e.g. cooling, presence of water tanks or playgrounds, construction materials) at growing scale will be automated with machine learning, trained with low human error by our labeling strategies. 

Youth and especially young women refugees are the most affected by unemployment crisis in the host countries. TaQadam solution is built on extensive human centered design research and interviews of Syrian and Palestinian refugees in Lebanon. Many named specific challenges of soft skills, language and payment that TaQadam solution targeted to solve. 

https://impakter.com/stories-invisible-refugees-tell-us-building-sustainable-solutions-challenging/

TaQadam solution is a result of human-centered design, which identified an income as the highest priority. As an ‘acupuncture point’ in the ecosystem of other players and solutions advancing human development, with our growth we ‘buy the time’ for youth to participate in partner-driven advanced IT training, incorporate e-learning, and ensure further entrepreneurship development and human security. We expect to bring our agents out of the cycle of systemic poverty by operating in the holistic programming ecosystem and ensure sustainable development and progress of each individual and their families. 

For the business, we will scan satellite imagery for a map of critical infrastructure. Applying Computer Vision leapfrogs remote sensing science. We will extract asset information from several sources of imagery matching it with other datasets. Reliable results generated by computer vision depend on high quality input data for machine learning, availability of high quality images but also that the labeled data is verified with consistency of judgement over features. TaQadam has experience in applying models augmented by continuous human-in-the loop labeling and verification, to build quality datasets for business. Taqadam uses its expertise, allowing our customers from public and private sector to enhance programming, and risk assessment. TaQadam will build a platform where the labels, imagery, classes, semantics, and formats are consistent for open-source.

TaQadam built web data collectors on locations for businesses; machine learning and human-verified asset datasets from satellite & airborne imagery for classes in agriculture and manufacturing. We were able to solve complex datasets challenges with 97% accuracy machine learning and features of
construction material, colors, shapes. 

Majority of tech-driven interventions to bring refugees to the digital economy focus on re-skilling and capacity building and often miss to address the challenge of missing infrastructure for digital jobs. Businesses in need of data annotation often resort to using crowdsourcing platforms, which require the crowd workers to have a bank account, speak English to navigate - therefore, inaccessible to the refugee youth. TaQadam brings an innovative crowd management through gamification. For client delivery, we create specialized teams, reward champions to ensure: timekeeping, domain expertise, and quality control. Our agents work remotely, on the mobile app, locking up low prices and high quality. 

TaQadam’s mobile-accessible gamified work and cloud architecture allow us to deliver data annotation work to clients at scale with high quality.  Our mobile app is an innovation for inclusion that brings access to digital work to refugee youth, which often faces a challenge of being unbanked, living in the areas of poor electricity and no access to computers, as well as having cultural barriers preventing them from independent freelancing and vocational work. 

Host conflict-affected economies cannot absorb and utilize the young talent leaving thousands of tech-savvy, educated youths working in exploitative informal markets despite hundreds of upskilling trainings. TaQadam solution is designed as an infrastructure connecting global businesses building data-powered systems for Artificial Intelligence and the refugee talent as imagery analysts/annotators. Working in TaQadam is a critical income generation opportunity, that buys the displaced youth the time for further professional development and learning, brings financial inclusion and builds social cohesion across communities.  We transform the displaced youth to a critical workforce advancing the future of technology.   

To date TaQadam onboarded 100+ agents: Syrian, Palestinian and Lebanese, in 5 different areas of Lebanon, and from Erbil in Iraq. Entire pool of 100 agents participated in our private beta client projects in April 2018. In July, we are starting the second pilot program retaining the original hires and build up ‘champion’ capacities. https://www.youtube.com/watch?v=k7WG7y5eetc&t=13s

We estimate to directly impact lives of 350 in the first year, and scale to 5000 and 8000 in the following three and five years respectively. With the average pay out rate of $10 dollars a day to an agent (capped at 2-3 hours a day), and poverty line in Lebanon at $3.8 a day, we anticipate to also impact our agents’ families and reduce poverty.

As a mission-driven enterprise, we are set for the journey with many more challenges than a regular AI tech start up, such as politics and populism, growing societal stigma against refugees, extra resources to set up the processes and operations in the conflict-affected countries, while competing in the very dynamic market of technology and AI. 

We believe the recognition by Solve will be transformative for our development and will bring us partners, business and funding opportunities, technology mentorship from MIT.

We will be able have the credibility and the exposure to show what a refugee with a smart phone can do. Which will allow us to have more costumers and thus have a sustainability. The funding and acknowledgment we get will help us establish better in Lebanon and register legally which will ease lost of financial operation fro us and for our youth workforce.

We are a 4 women in the solution team. 3 of them are refugees and serve in roles of:

1. Karina Grosheva, CEO

2. CTO and Client Success,  Nawal Messi

3. Sales and Operations,  Sendus Majanni

4. Marketing and Lead generation, Shirin Sobhani

We share a passion to advance TaQadam’s mission and bring a combination of skills in business, technology, AI and social impact to deliver on the challenge. All of us share a story of living and experiencing many of the challenges that migrants, internally displaced and refugees face. 

As a result of previously participating in Peacetech Accelerator we partner with:

- AWS and Peacetech labs

Last year we were winner of Refugee innovation at Techfugees, and we partner for advocacy and technical outreach with them.

https://techfugees.com/projects/techfugees-x-taqadam-to-provide-remote-work-in-ai-to-refugees-in-the-middle-east/

We also won Touch competition in Lebanon, and partner with them to expand Taqadam in Middle East: https://www.youtube.com/watch?v=8_Z1VpdqVg8

We are currently exploring partnerships with Digital Globe and Planet labs as imagery providers for our projects related to aerial imagery annotations.

TaQadam business model is built on agility and flexibility to match clients’ data processes. We charge companies for  annotations - from $0.05 per object detection to $0.15 dollar for polygon segmentation for feature detection.

With seamless technology  of API and cloud architecture in place, we reduce client acquisition costs, build on agents’ learning curve and enhance quality/speed of annotation service with machine learning.

Our go-to-market strategy includes AI implementing startups, commodity AI as well data-driven companies from emerging markets, including from the Arab world with high aspirations for AI implementation. With the right traction, we aspire to serve the race for AI domination, and scale fast and sound to meet the growing global demand.

We also, explore providing full direct service on alternative data with acquisition of the satellite imagery from providers.

From robots, drones to satellite imagery, the market need for quality data for artificial intelligence is extraordinary. For AI enterprise solutions, what really matters is the quality of data for training models. Today’s data is driving tomorrow’s AI products. To be competitive in AI, innovation depends on having data-edge often more than a technology-edge, but 80% of data engineers’ time is spent on data work. TaQadam optimizes image annotation for specific use cases and serves company to unlock imagery for insights - prospecting, understanding assets and changes. With an API we ensure a simple and secure way to build image data set with high accuracy and precision. With gamification and mobile accessible work, we bring opportunity even to the most vulnerable youth. By leveraging public sector resources and UN-led ICT training, we extend work of the future to the youth in the post-conflict zones and ensure scalability, sustainability and high level of retention rates. 

We live during the era of great turbulence: political paralysis and fragile economies, deep inequities and inadequate systems for generation-Y and Z, and a vastly more interconnected and independent world. Problem-solving for the required transformative changes is not to be delivered alone. We believe in a power of community and networks - to learn, grow and innovate for the most pressing wicked problems with technology. In TaQadam we consider ourselves as ‘political entrepreneurs’, solving a complex phenomenon of ‘human rights poverty’ in the post-conflict economies, and seek for a visibility, support and advocacy for our cause.

By recruiting from partner institutions, we extend work of the future to the youth in the post-conflict zones and ensure scalability, sustainability and high level of retention rates. To date TaQadam connected with Lebanon UN livelihoods cluster consisting of 49 organizations. We partnered with NGOs and UN agencies rendering ICT training to help us recruit, vet and onboard to TaQadam. 100+ youth joined TaQadam: Syrian, Palestinian and Lebanese, in 5 different areas of Lebanon. Entire pool of 100 youth participated in our private beta client projects in April 2018. In August, we are starting the second pilot program retaining the original hires and build up ‘champion’ capacities. https://www.youtube.com/watch?v=k7WG7y5eetc&t=13s

We are currently exploring partnership with geospatial imagery providers: DigitalGlobe (Maxar technologies) and Planet labs to accelerate the development of service related to aerial imagery and asset/infrastructure mapping.

We are advocate of AI for good. Our solution allows companies to train data to apply Artificial intelligence in both industrial automation as well as using drones and airborne imagery collected for sustainable development. To date we developed our own ML models to allow:

- bring innovation to farming and agriculture sector by better understanding and mapping assets and supply chain

- understand industrial assets

- bring productivity and safety by training sensors detecting anomalies on production sites.

We envision that we are unique to offer both job creation as well as AI innovation to companies, cutting through the bias, and opportunity of AI future to impact the vulnerable communities as well.

We are excited to participate in GM prize as we see companies expanding their innovation portfolio with our solution, while doing good and impacting communities.

We envision that GM prize will allow us to develop a deeper research and focus on business development for the specific of the industry.

Majority of our annotating team is women. 

Our solution team consists of women only.

We are interested to power up gender equality in technology and will use the prize to focus on technology training across our team, and strengthen our internal capacity to run AI-startup.

TaQadam's core value is refugee inclusion. We focus on the community of young individuals that often lost their educational credentials, but within the year of working with us proved a value of growing into AI-analysts and being fully empowered through the part time employment.

The future is in data quality. Annotators are programmers of the future. We believe that our value is higher than temporary income generation opportunities. It is refugees, working in Lebanon and Iraq, bring value to the companies.

Complex global systems of physical assets in agriculture, resources, manufacturing, infrastructure, and government services affect billions but are still poorly understood, despite years of remote sensing data collection. Understanding the geo-coded asset maps is critical for both sustainable development, as well as for businesses that adapt supply chains in the fast-changing environment of our planet. 

With this prize, we will scan satellite imagery for a  map of critical infrastructure assets. Applying Computer Vision leapfrogs remote sensing science. We will extract asset information from several sources of
imagery. Reliable results generated by computer vision depend on
high quality input data for machine learning, availability of high quality images but also that the labeled data is verified with consistency of judgement over features. TaQadam has experience in applying models augmented by continuous human-in-the loop labeling and verification, to build quality datasets. Taqadam will use this previous expertise for this project, allowing data users to enhance programming, and risk assessment in a number of critical programs. TaQadam will build a platform
where the labels, imagery, classes, semantics, and formats are consistent for open-source.

TaQadam

TaQadam