RelationalAI Year in Review, 2022

As the year draws to a close, we’re taking a look back over 2022 to gather all our best content from the year into one place. It was a great year for us — we came out of stealth, grew our team, participated in fantastic conferences and events, and we’re excited for everything 2023 will bring.

Thank you for reading our blog this year and keeping up with our news. We wish you a very happy new year from all of us at RelationalAI.

We started a newsletter this year — sign up to get all our news straight to your inbox once a month. No spam; just industry news, insights, and updates.

Quarter 1 — AI-Enhanced Customer Experiences

A Fortune 50 retailer drove $1 billion in incremental revenues over three years after deploying AI models developed by RelationalAl. Learn how relational knowledge graphs are used to create AI-enhanced customer experiences.
We celebrated International Pi Day on March 14 with a fun blog post demonstrating how π can be approximated in various ways in our declarative modeling language Rel.
Taking inspiration from relational SQL databases, graph database research has achieved breakthroughs in algorithms and storage technology. We explored the false dichotomy of ”Graph vs. Relational” to show how RelationalAI puts it all together.

Quarter 2 — Our First Press Release

In May we announced that we raised a total of $122 million in funding after closing our Series B round. We are using the funds to accelerate product development and go-to-market activities for our relational knowledge graph system. With thanks to our investors Tiger Global, Madrona, Addition, and Menlo Ventures.
We were honored to be named a Gartner Cool Vendor in Augmented Data Management. We are delighted that the impact of bringing data application, graph, and machine learning workloads into a single relational system has been recognized.
We participated in the Knowledge Graph Conference in New York where our board member Bob Muglia presented the keynote speech. Bob discussed the growth and coming of age of the modern data stack and the role he sees for knowledge graphs with a semantic layer.

Quarter 3 — Promoting City Resilience With the UN

The International Conference on Learning Representations (ICLR) 2022 was full of exciting papers. We spent roughly 100 hours going through the content, as we believe this research will drive the commercialization of AI. We presented what we found to be the most noteworthy ideas in the areas of language models, self supervision, sparsity, and application in sciences and fundamentals.
The International Conference on Machine Learning (ICML) 2022 took place two months after ICLR 2022, and we dived into the research to present a summary of the trends in machine learning. We highlighted cost reduction, Machine-Learning-as-a-Service (MLaS), research into biomedical AI fueled by the pandemic, and the return of Vowpal Wabbit.
We partnered with the UN Office for Information Technology to support the analysis of ongoing stresses and sudden shocks in cities. In this presentation at Ai4 2022, we showed (opens in a new tab) how combining graph analytics, simulation, and reasoning within a single system is used to understand how disasters impact cities.

Quarter 4 — The Magic in Graph Analytics

We celebrated Halloween on October 31 with a logic puzzle in our declarative modeling language, Rel. Come trick-or-treating with us any day of the year as you learn how to model a problem, store facts, and infer new knowledge.
Our board member Bob Muglia recently met with Sanjeev Mohan in an interview for the It Depends podcast. Bob and Sanjeev discussed the challenges, trends, technologies, and the general pulse of the ever-changing data analytics market.
Graph analytics help us make sense of our connected data by understanding the structure of our data. They help us see which patterns are important and which aren’t. They help us predict what’s coming next. We explored what makes graph analytics feel so magical and how they differ from typical analytics.

It’s been a busy and exciting year for us! If you’re as excited as we are, we’d love to hear from you. We have career openings on our Field, Product, and Development teams. Have a wonderful 2023!

Building a Named Entity Recognition Model for the Legal Domain

We defined NER in the legal domain and presented our approach towards generating ground truth data. In what follows, we go over the state-of-the-art in the NER domain and elaborate on the experiments we ran and the lessons we learned.

Named Entity Recognition in the Legal Domain

Named entity recognition is a difficult challenge to solve, particularly in the legal domain. Extracting ground truth labels from long, hierarchical documents is often slow and prone to error. RelationalAI proposes a new scalable algorithm based on the principles of data-centric AI, designed to meet this challenge and generate high-quality annotations with minimal supervision.