Riskboard

The gap between organizations and the communities affected by their supply chains is too large. Monitoring complex supply chains for human rights abuses is challenging across many product lines and markets. Riskboard aggregates and monitors signals using traditional media, social media, and custom databases in multiple languages to detect communities facing human rights issues and maps it against supply chains to help CSR / risk teams and activists improve supply chain integrity. Beyond detection, RiskBoard provides analysis of the entities involved, history, and sentiment scores to help users understand the event's significance. We are like a Bloomberg terminal for human rights and ESG. We’re already designing and prototyping with Greenpeace and New Balance, but scaling globally to Russell 1000 companies, ESG investors, and human rights advocacy groups will help us amplify the voices of the vulnerable as corporate governance rises in importance and is mandated by law.

The gap between organizations and the communities their supply chains affect is too large. As a result, human rights abuses continue to persist in supply chains with limited ability for global enterprises or advocacy groups to detect and highlight them. According to the annual political risk survey by Willis Towers Watson, 55 percent of companies worth at least $1 billion have suffered a loss based on political risk in 2018.

A major retail chain told us it was impossible for their small team of analysts to monitor their 30,000+ product lines across 100+ countries for human rights or political risks, often resulting in their risk and CSR teams failing to recognize and respond to human rights issues. We've spoken to over 50 other organizations who share this challenge. 

The more complex the supply chain or investment portfolio, the harder it is to hear the voices of local communities downstream. There are few tools to bridge this gap between community voices and supply chain integrity and ESG teams. With the abundance of modern data, Natural Language Processing, and other machine learning techniques, this problem is now solvable in real-time.

First and foremost, we are serving vulnerable communities and workers in the supply chains of manufacturing, resources, and primary industries by highlighting issues that affect them. We put their concerns on the map by deploying cutting edge analytics and machine learning to detect and predict potential human rights abuses related to supply chains. Our second constituents are (1) global Russell 1000 companies with complex supply chains and human rights reporting obligations e.g. New Balance, and (2) global human rights advocacy groups e.g. Greenpeace, who can hold parties to account by highlighting issues as they arise. We are already working with Greenpeace and New Balance as design partners and several other potential clients who we can scale to through the Harvard Corporate Social Responsibility Initiative.

We provide real-time human rights and political risk detection (unlike static consultancies) with personalization mapped against supply chains. Our tool also monitors more data sources, geographies, and languages than human consultants could. Many of our data sources are also private. 

RiskBoard automates the monitoring of traditional media, social media, and custom databases in multiple languages to detect political events and stories that could directly impact an organization’s international interests. Beyond detection, RiskBoard provides analysis on the actors involved, history, and sentiment scores to help users understand the risk’s significance. It also offers examples of past attempts to mitigate risk, which can empower organizations to proactively manage their risk.

Our advantage is that we provide real-time political risk detection (unlike consultancies that are static) with analysis that’s personalization to the client’s assets or supply chain (unlike alert tools which offer general alerts). The quality of our analysis will also outperform consultancies in some instances as our tool monitors more data sources, geographies, and languages than human consultants could hope to cover and draws links that analysts may not be able to observe. By working with Greenpeace and other organizations working on the ground, we’re also able to develop valuable datasets that are not open to the public, particularly corporations. Finally, our experience in machine learning and global political risk sets us apart from most in the industry who have either studied one geography in great detail or the world of media alerts, without bridging this gap to provide meaningful, tailored insights and empower organizations to proactively manage their political risk.

First, our solution is significantly more affordable than the traditional human rights due diligence consultants, the primary actors in the space. Our lower cost reduces the barrier of entry for companies to start monitoring human rights within their supply chain. Additionally, our solution has several qualitative advantages: instead of a static, annual report, it provides real time insights and analysis allowing rapid response. 

Furthermore, there is a gap in the market even when comparing RiskBoard to the other technology-enabled competitors. The competitors above: do not monitor open source information to identify relevant threats as they emerge, only serve as a way to crowdsource feedback from workers (in forced labor relationships, workers often face coercive forces to not utilize these tools) or only look at a company’s supply chain, rather than accounting for external political / economic forces that can create risk. We plan to test our assumptions by piloting our product with a large company to ensure companies benefit from the product. We also plan to speak with large companies that have utilized our competitors’ services to further validate and understand pain points.

As mentioned we have developed a software-as-a-service platform that utilizes large amounts of aggregated data. This data includes local news sources, social media sentiment analysis (both of which are translated and processed using a natural language processing), and political / economic indices. This data, in combination with a company’s internal data, is analyzed using a machine learning algorithm we have developed (through the team’s political risk analysis expertise) to create a forced labor risk score. Based on the specific nature / level of risk, the artificial intelligence platform provides a recommended course of action (e.g., audits, shutdowns, etc.), drawing on a large database of company responses to past incidents. We are also in the process of figuring out how to aggregate, store, and analyze historical data, so that we can develop a machine learning algorithm that can utilize this set of information to predict forced labor violations before they occur. Longer-term, we plan to try and expand beyond forced labor also including other human rights risks (e.g., child labor, unfair labor practices, environmental sustainability challenges). 

The fundamental problem we are addressing is forced labor in supply chains. We made several assumptions in dealing with this. First, we assumed that companies care about this and the problem we are solving for is that it is difficult for companies to monitor their value chains. We have validated this by speaking to over 50 risk, compliance, and procurement officers at several large companies, human rights academics, and company board members. These interviews indicated that this is a fundamental concern for companies, who are worried about regulatory compliance, reputation, consumer social-mindedness, and promoting human rights.These interviews highlighted that companies have difficulties governing suppliers through vendor codes of conducts because of a complex tiered system and the use of off-book subcontractors. 

The second assumption is that we can actually predict human rights risk based on open source information. To test this we have built a prototype of the dashboard based on our knowledge and the customer and academic inputs we have received. This prototype is focused on Thailand’s seafood industry (known for having a high risk of forced labor). We have been able to effectively process and analyze inputs from news sources, social media sites, and political and economic indicators. The prototype has also been able to identify red flags that indicate human rights abuses before they received international news coverage. We are in the process of building: linkages to internal client data, additional natural language processing capabilities, and a machine learning algorithm that can identify trends forecasts risks.

We are currently still prototyping with our design partners but hope to work with Russell 1000 companies to serve the workers and communities affected by their supply chain that would number in the hundreds of thousands.

We plan to complete our prototype testing by September in coordination with our design partners. We then plan a soft launch in October before scaling and launching with full force in 2020 through a coordinate sales effort. 

Our primary barriers at the moment relate to engineering horse power. We are progressing with the build but will need to add additional engineering muscle to work on data engineering on the back end. We will then need to focus more on user experience and design, for which we will also need expertise. 

We plan to raise pre-seed capital to allow us to hire additional engineering staff. 

Four:

Arjun Bisen, Harvard Kennedy School, Public Policy and Tech

Ryan Jiang, Harvard College, Computer Science
Pradeepan Parthiban, Harvard Extension School, Human Rights
Ori Pleban, Harvard Kennedy School, Data for Government

Our four-person Harvard team consists of a diplomat and technologist, an entrepreneur who has worked on space and hyperloop startups, and data scientists. This gives us specific strengths in the subject matter as well as the implementation of building the front and back-end of a data-intensive tool. Our team also has direct links to Harvard's Carr Center for Human Rights and the MIT Media Lab, both of which provide access to key advisors and experts in this field. We have mentors from MIT Sandbox and the Corporate Social Responsibility Project at HKS. We also have Professor Satchit Balsari and Professor Tarun Khanna as advisors as our project was developed through their class.

We are working with Greenpeace and New Balance as design partners for the product and working with the Mittal Institute of South Asia, The Harvard Corporate Social Responsibility Initiative, and several Indian NGOs to build a stronger footprint and network on the group to compliment our analytics tools. 

We are a enterprize SAAS company that will charge a subscription fee. 

2 Subscription Tiers - $20K/year (general) and $75K/year (custom)

Breakeven in Year 3

Average Revenue Per Customer: $38,333

Projected Annual Revenue of $1.9 M by 2023

We plan to bootstrap the company with a Customer Lifetime Value of $35,714 over 40 years

LTV/CAC ratio: 4

Gross Margin:

36.50>

Solve can help us recruit talented engineers passionate about social causes. It can also help us connect with more potential customers and think about new markets we might expand into.

We think partnering with the IOM and UNOHCHR could help us scale our products and access unique data that could improve our product.

RiskBoard automates the monitoring of traditional media, social media, and custom databases in multiple languages through NLP and AI to detect political events and stories that could directly impact an organization’s supply chain. Beyond detection, RiskBoard provides automated analysis on the actors involved, history, and sentiment scores to help users understand the risk’s significance. It also offers examples of past attempts to mitigate risk, which can empower organizations to proactively mitigate human rights risks.

We will use the prize to hone our machine learning capabilities and provide real-time political risk detection (unlike consultancies that are static) with analysis that’s personalized to an organizations supply chain (unlike alert tools which offer general alerts). The quality of our analysis will also outperform consultancies in some instances as our tool monitors more data sources, geographies, and languages than human consultants could hope to cover and draws links that analysts may not be able to observe.