Tensoriel

We believe corporations, big or small, have to become global partners with local communities on many sustainability and social issues to have a real impact.  But news regarding corporate actions impacting local environments often get generated by companies themselves, a process sometimes called “greenwashing”.   A solution would be for stakeholders to be aware of most professional, local, and “hyperlocal” news originating from both professional publications and community organizations.   But that would require an immense journalistic staff.   We use Artificial Intelligence and Natural Language Processing to process and categorize about 200,000 articles per day from more than 70,000 different premium sources from all over the world.   Using our tools, readers can then select industries, companies, and topics they care about such as “Carbon Emissions”, “Water Management”, “Human Rights”, “Diversity and Inclusion”, etc.  Our users can then check and assess the real impact of corporate actions on sustainability and social issues. 

Many articles about sustainability issues are published globally.   They address many topics in many languages in many journals and publications.  It is a big challenge for decision makers to read and process this huge quantity of information.   We built a framework using Artificial Intelligence and Natural Language Processing, to process and categorize articles in real time.   We use more than 70,000 sources and process more than 200,000 articles daily, related to about 200,000 companies worldwide.  Our users can then select industries, corporations, and topics to follow a supply chain, impacts on water management, inclusion and diversity, human rights, etc.  With our technology, we can adapt to many different needs across industries and geographies.   For example, we worked with a Transportation consortium in Europe.   This organization wants to be aware of the many changes happening in their industry including autonomous vehicles, alternative fuels, fair access to public transportation, etc.  For this consortium, we added 700 small companies and 100s of specialized sources to our databases.   As many articles are “hyperlocal”, they usually do not make it to the top of relevant news.   Our solution provides a louder voice to local communities to expose concerns that are relevant to them.  

Our solution uses state-of-the-art AI and NLP technology to process large amounts of articles published all over the world.  Our processing pipeline is complex and runs “on the cloud” in real time.   For example, one of the steps is called “Named Entity Recognition”, and makes it possible to detect what corporation, university, or government organization the article is talking about.  We use various statistical tools to detect salience, check on outliers in times series, use clustering analytics to detect and gather specific events, etc.  The entire pipeline runs constantly, as premium publications or social media publish local or global news.  Our technical team has expertise in AI, Machine Learning, NLP, and NLG (Natural Language Generation).   We use MongoDb as our primary data store and various specialized technologies from Google and Amazon to make our processes scalable and able to process constantly increasing amounts of information.  We design linguistic rules to match existing sustainability frameworks (such as the Sustainable Development Goals from the United Nations).  In our basic offering, it is already possible for communities to create and customize multiple “feeds” that include a selection of companies, topics, languages, locations, and settings for daily digests and alerts.     

Our solution serves both corporations and communities.  We process about 70,000 premium sources of information globally.   When we work with a specific industry (e.g., Transportation), we integrate “hyperlocal” sources related to the target industry.  These sources are from local blogs, newspapers, small businesses, or from communities, at all levels, cities or smaller.  News from these sources tend not to make it to the top of major newspapers or magazines.   Our technology makes it possible to give them a voice equal to the large publishing organizations.   Typically, corporations who want to know how they impact local communities may engage into a long, complicated, often biased, and expensive audits.   We believe that when corporations can get immediate feedback from communities (some of our topics include Human Rights, Labor Practices, Hiring, Water Management, etc.), they will tend to improve their practices more quickly and more efficiently, and will improve lives of local communities.   Our process tends to encourage cooperation and discussion through all stakeholders, and tends to avoids conflict that emerge from lack of specific and concrete information.   As we address more and more industries, and aggregate more and more voices, our process will also scale up in engaging all stakeholders.  Recently, we integrated a state-of-the-art technology usually called "no-code".  Using a sophisticated API, communities can select sources, topics, language patterns, locations, that match their specific needs.  Community of users, within a single company, a consortium, a city, or any organization, can be seen as training an AI to help them achieve their goals and avoid biases, and making the resulting AI part of their own team.  

We actually address all Challenge dimensions.  We selected one as we have to select one.  But as mentioned above, we process and categorize huge and varied quantities of information through our AI technology.   Topics include "Diversity and Inclusion", "Ecological Impact", "Water Management", "Labor Practices", etc.   We target all of the 17 United Nations SDGs.  We monitor 200,000 companies and how their actions may impact these topics in real time.   We also customize our process to specific industries (e.g., Transportation).  Our goal is to make such information available to both corporations and communities or any size.

We delivered our Solution to a Transportation Consortium based in France.  This consortium gather many transportation companies, large and small, that want to keep informed about how the transportation industry is currently evolving.  Communities include small business owners and employees who are members of supply chains to large companies.   Through a partnership, the consortium provided us with keywords (e.g., biofuel, autonomous vehicles, etc.), small hyperlocal sources, and specific topics (e.g., school transportation).  We then integrated their specific needs to our generic processing system.  Currently, only 6 experts of the consortium are using our system daily, but it is the role of the consortium staff to broadcast the information we provide to all their members and to their communities, which consists of 1000s of workers.   The integration technology we built can be extended to any specific industry all over the world, giving a voice to many small communities.

Our solution uses recent advances in Artificial Intelligence and the cheap computing power now available in various clouds.  Our solution can apply to any industry or company in the world.   We provide both a generic solution and an Enterprise solution.  Through the Enterprise solution, we partner with domain experts and give voices to small publications and news sources.  We process 300,000 articles daily, and let human experts focus on the top of the information supply chain.  Within our solutions, we can provide multiple editions.  Our current focus is a "Sustainability Edition".  News processing and topic categorization is a first step.  Following progress in AI, we intend to focus on analytics, sentiment computation, risk evaluation, impact assessment, etc.  As our solution can be used by any business, we intend to change the market.   We plan to be the "Google" of sustainability for the business world!

We use innovative mixtures of deterministic rule-based models and statistical models (machine learning, non-linear regression).  These two types of models have known benefits and drawbacks and can complement each other.   Using Bayesian statistics, we created our own formal mathematical models, such as a "salience" model or a "sentiment" model, to keep results consistent across models and to generate optimal mixtures.   Rule-based systems and stochastic models are hot topics in AI that we will contribute to.  

In March 2021, our company was included in the list of the best and most innovative startups in France by the business magazine "Challenges".  

The previous questions are "In which countries do you currently operate?".  As a SaaS and DaaS provider, we are global.  We use a database of more than 200,000 companies worldwide and we process localizations of every country, state, region, city, in the world.  As long as users can speak English, or French (more later) we can operate in any country.  It is important to us to detect specific locations mentioned by each article we process.

Through the Transportation consortium, we serve 5 transportation experts (with focus in energy, mobility, urban, rural, access).  Through them, we serve all companies in the consortium and their employees (several 1000s), and then indirectly all transportation users (a large number...).   We provide missing links between scientific progress on fuels, energy management, clean air, health, and the transportation industry.  

Within a year, we plan to serve any company involved in sustainability and impact: Companies, large or small, asset managers, private equity, journalists, funds, philanthropies, consultants, you!

Within 5 years, we plan to respond to any demand to any industry.  We believe AI makes it possible for us to scale up to such demands without incurring large expenses.   As an example, one of our main prospects currently works for the agriculture industry and is interested in how technological progress can provide solutions to hunger in under-developed countries (UN SDG 2, WB 332).

We provide SaaS or DaaS on the cloud.   We can measure progress in different dimensions:

- number of sources.   Currently, about 50,000 sources.   We recently mapped several 1000s sources that focus on the 17 SDGs and on SASB.  Our system can quickly add 100s or 1000s of sources if needed.

- impact taxonomies.   We use the World Bank taxonomy and map their terms to SDGs.  The WB taxonomy is very detailed and pragmatic.  As the UNs are more "country focused" and the WB is more "finance focused", we can also focus on global, pragmatic solutions.  We believe SDGs are successful because they were built from conversations with many stakeholders and because they include targets, indicators, and time series of progress.  We believe this is the correct approach and that our technology can bring substantial contributions without incurring huge costs.

- impact measures.  We are fully aware that industries are demanding "standards" in impact measurement.  A number of organizations are moving in that direction (e.g., OECD).   A lot of manual labor is still needed to compute impact, from manual entries and from many rows and columns spread across multiple spreadsheets.  As mentioned above, automation is one of our strengths: our experience with the financial industry and quantitative asset management, our knowledge of data mining and modeling, contribute to our vision on how to measure impact in the near future.

Full time: 4

Part-time: 2

Contractors: 1

Interns: from 1 to 3, from our academic relationships.

Our team has exceptional skills and knowledge in data, technology, and finance.  Our goal is to use AI to provide data and analytics to any user interested in sustainability, impact investing, ESG, UN SDGs, etc.

We are a small global team, located in France, Morocco, and the United States.  We originate from Europe, Africa, Asia, with team members who migrated across continents.  

We all went through difficult times during our education and later.   Some of us have known hunger and were almost homeless at times.  We have acquired skills (PhD, Masters, CFAs) through grit - and luck!  We know risk.  We know how difficult it can be to get good education when you are "on your own".  We have known poverty, and we have a passion to use our knowledge to help others.  Three of us spent time or are still living in Africa.  We have hired an intern (female gender) from Ghana.  We understand how diversity of staff and opinions can create value.  As mentioned above, our system can provide voices to communities that have none, independently of size or location.  We want to learn from any new community, as we have learned from our own communities during hard times.

At the same time, we understand business, we have worked for both small and large corporations, we are pragmatic, we understand data, we want to solve problems, connect to businesses, and see how we can help them to be both profitable and helpful. 

Our CEO originates from Morocco.   He has immense talent in math and technology.   He was able to join the best schools in France and in the United Kingdom, and later to attend Stanford University.

Our CTO originates from and lives in France.  He also teaching technology at the University of Lille.  His is young with exceptional tech skills.  He also has a passion for teaching.

Our Head of Sustainability Intelligence is originally from France.  He emigrated to the US to get a PhD from the University of California.  He is a pioneer in AI and Machine Learning.  He published a book and about 50 articles in peer-reviewed tech publications.  He started to work on ESG and sustainability in the 90s.  He worked on "Life Cycle Analysis" and built a tool called "TEAM: Tools for Environmental Analysis and Management", later sold to Accenture.

Our Head of Data is originally from Hong Kong.  She also emigrated to the US with her family.  She has a long experience (more than 20 years) in investment management, and in data and analytics.

Our team reflects equity and inclusion because we are ourselves the results of equity and inclusion.

As mentioned above, one of our goals was to make it possible for businesses or communities to focus on a ESG topic or a SDG goal (e.g., SDG 5 or 10).  The sustainability edition of our SaaS product is a solid foundation that can be "fine tuned" to specific needs in any domain using "no coding" API.

Our main barrier is visibility, marketing, and communication or connectivity.  We used to work on AI and Life Cycle Analysis in the 90s, well before AI and sustainability were in the news.   Sustainability has made progress and AI is making progress.   But there is still misunderstanding regarding what AI is about, especially with language understanding or generation.  

We heard of "Solve" at a conference.   We were impressed.  We checked you out!  We noticed you were global, interested in SDGs and ESG, and you were helping a few teams involved with AI.  We immediately thought there was a match.   We would need Solve to open doors with any corporation, community, or asset manager interested in knowing what AI can do to help with SDGs, to bridge a gap between complex scientific data and shallow social media.  We are at the first step of a long journey.  We believe Solve can help us with the next steps. 

Solve is connected to many organizations, communities and businesses.   We need "entry doors", in a way similar to what an incubator can do.  If we get a first interest, we can pitch and demonstrate our product.  Global communication products built by zoom or Google make it easy for us to connect globally with little cost.  The uniqueness and the power of our product then become quite convincing.

Business Model.  We do spend time and effort to understand needs for services like we provide.  But as explained above, our potential market is huge.   We have to focus where business results are larger and more probable.  Also, for example what if we give 10% of profits or revenues to organizations that work on cleaning the oceans (SDG 14)?  Will that help?  Is that a good strategy? 

Financial.  We believe we can pitch well to investors.  But we need entry doors and a good selection of potential investors with good reputation.

Public Relations.  As explained above, we need help with Public Relations ("what can AI do for you and your sustainability needs?").  Again, we believe that if we can provide a first pitch to prospects, we have convincing arguments our services can help them quickly and for a low cost.  We need Solve to put us in touch with decision makers that are willing to listen to us.  We use LinkedIn, we are making progress with MailChimp and Hubspot, but we need access to global media.  Using Artificial Intelligence to help with climate change, for example, could become a powerful message under a proper strategy.

Product/Service Distribution.  This is the crucial category.  SaaS distribution is simple.  Through the cloud, we have elastic computational power that can adapt to demand.  But expanding the client base requires more effort, as we need connections and time from busy professionals, who often have heavy responsibilities.

Any organization that can "open their first door" for us would be a good partner.   We believe Solve has the knowledge, skills, and contacts that would make it possible for us to explain and show what we can do.  For each possible partner at Solve, we could jointly target an asset manager, a corporation, a philanthropy, a fund, who would be willing to listen to us and to share our vision and enthusiasm.  From a first meeting, we are confident will be get a high positive hit rate of success.

Of course, MIT faculty has a superb reputation worldwide, both for technology and business know-how.  We would be quite honored to consider a faculty member as a partner, or possibly as a board member.  As previously mentioned, our own connections are with London or Paris academic institutions, or with the University of California or Stanford University in the US.  

We do not have specific names in mind at MIT or at Solve.  We believe the judges of our solution would also be the best judges of names who could become the best partners or collaborators with us.  We are certainly willing to have further discussions with you on this topic as we recognize the value of such possible partners.

(Could there be an issue with this question?  The question addresses Refugee Inclusion which seems to be related to the Andan Prize, not the GM Prize).

Regarding the GM prize, as mentioned, we focus on sustainability and SDGs, which include SDGs 4 and 5 on Gender Equality and Quality Education. Here is a short list of the topics related to education we address in our platform: Access to Education, Education Financing, Science and Technology, Teachers, Student Assessment, Education Governance, School-Based Management, Curriculum, Education Facilities, Private and Public Education.

Considering our specific SaaS edition on Transportation, we would find it quite interesting and valuable to link the social goals of General Motors to these specific SDGs!  GM would also be able to compare the results of their initiatives to any other company in the car industry, or within their supply chains.

On top, our global approach can address education topics in the United States  or in any country in the world, big of small.   

We use our AI technology to integrate UN SDG 5 with the World Bank taxonomy.   During that process, we included many topics relevant to the quality of life of women and girls.  They include job creation, job quality, entrepreneurship, income equality, financial inclusion, gender based violence, disability, conflict prevention, education, healthcare, nutrition, etc.

It is clear to us that any social or financial issue related to sustainability is much worse for women.  Our solution can show what companies may have a positive or negative impact in any region of the world.  Having a clear, daily information on the nature, the location, and the impact of gender inequality, is a first step towards a solution.

As mentioned, we target all SDGs.  We are especially interested in SDG 14.  In our generic service edition, we process many professional news sources that target "Life under Water", like "Mission Blue" or "Sustainable Fisheries".  Our AI can detect any organizations that have strong relationships with the oceans, positive or negative.   That includes overfishing, coral destruction, ocean temperature, biodiversity, coastal management, fisheries management, etc.   

As mentioned above, we have the means to build specific editions of our platforms (like we did for transportation).   We would then build such a specific edition.  Any corporation or investment fund would then have a detailed knowledge of the status of oceans and overfishing across the planet: What companies are involved, positively or negatively?  How communities are impacted, positively or negatively?

(By personal interest, and because of our upbringings, with a mixture of connections to the Atlantic ocean, Pacific Ocean, the Mediterranean sea, the South China sea, our team is especially interested in Life under water.  We volunteer, we sail, we dive, we surf, we swim with turtles, we feel connected to the oceans.  Recently, sailing down the coast of California down to Mexico gave us one more lesson about the cross-dependencies between fishing and local communities.)

From previous comments, it is clear we have a good connection with ServiceNow.  As we also worked on SDG 8, we could also build a specific edition of our platform dedicated to labor issues, child labor, forced labor, rural and urban work, etc.  

We integrated all labor related issues from the World Bank into our platform.  We could fine tune WB taxonomy into more difficult issues like immigration, human trafficking, gender violence, peace and justice, for any specific community in the world.   Our geolocation database includes countries, states, and cities worldwide. 

Carbon absorption is one of the topics we address in our SaaS platform.  We have spent specific time and effort on the "CleanTech" industry to categorize topics addressed by companies that specifically provide technology to improve air or water quality, and prevent global warming.  This industry is quickly evolving and makes our AI approach even more relevant. 

We feel there is a great match between our vision and this prize.   Our business is all about using AI to solve sustainability issues.   We are fully aware of what AI and Machine Learning can be good at.  We process much more information than a large team of human experts can do, daily, and globally.  

At the same time, we are also aware that this is not an easy journey.  We are at step 1, maybe 2!  We have been lucky to work under or to collaborate with exceptional scientists: Francis Crick, Nobel Prize for the discovery of the structure of DNA; Amos Tversky, partner of Daniel Kahneman, Nobel Prize in economics; David Rumelhart, one of the top pioneers in AI and Stanford Professor; Barr Rosenberg, founder of the Finance department at UC Berkeley.  We know science is hard.  We were trained by giants!

We consider language to be a large cognitive functions of what makes us human, and that AI/NLP, if well used and well understood, can help us as humans.  We are scientists who want to use AI science to make progress for the common good through pragmatic business applications.  We admire what Patrick J. McGovern has been able to do in this direction.  Obviously, we would be honored to be considered for this prize.