Saturday 28th August marked Seldon’s seventh birthday since incorporation in 2014, and we’ve achieved some fantastic milestones in that time. In the early days, MLOps was a nascent category. Now it’s rapidly becoming an essential part of IT infrastructure – not just to deploy models at scale, but to run models at scale and monitor model performance in real-time. Over the last few years, we’ve expanded rapidly in both the UK and internationally and have doubled our headcount since this time last year. There has never been a better time to be part of the Seldon journey, and our unique perspective and position in this emerging market mean our incredible team and community can look forward to shaping the future of MLOps for years to come. So join us as we take a look through some of the highlights along the way.
The Early Years
- Our first open-source ML platform Seldon Server, launched in 2015, was funded by angel investment and data science consulting projects. We originally intended to commercialise some of these “use cases” since the market understood this, but we soon realised this business model would be harder to scale. So we had to pick between building horizontal infrastructure or a verticalised full-stack AI company. The latter was a much better-understood proposition at the time, but we were more passionate about infrastructure. We believed that’s where the big and meaningful problems were and that we were the right team to solve these.
- In 2016, Seldon participated in the Barclays Accelerator, powered by Techstars. Through this experience speaking to hundreds of stakeholders in financial services, we saw first-hand how the model governance challenges were a limiting factor in the adoption of machine learning. So we went back to the drawing board to build the next generation ML deployment platform. Pitching at the Demo Day at London O2 was an unforgettable memory and helped to sharpen the company narrative.
- In 2018, we launched our open-source model-serving platform Seldon Core. This product is now utilised by tens of thousands of organisations across the globe, with more than 1.5 million machine learning models deployed and more than 100 code contributors.
- Model monitoring and explainability are essential parts of managing the risk of running machine learning in production. In 2019, we launched the Alibi Explain and Alibi Detect python libraries which provide a collection of cutting-edge algorithms engineered for performance and scalability. Through building Alibi, our data science team has published four well-cited research papers, some of which have been accepted by the JMLR journal, which is considered as the top journal in AI/ML academic fields, and leading conferences including NVIDIA GTC and ECML PKDD 2021 – Alibi even features in some textbooks including “Explainable AI with Python” (Manning).
- Our open-source projects continue to make it easier and faster to deploy, monitor and explain machine learning models than ever before.
- Over the past couple of years, Seldon has raised more than $13 million in venture funding, which has helped us rapidly expand our team of researchers and engineers, pushing forward our state-of-the-art products. We’ve hired people who love building things and a commercial team committed to helping our customers leverage MLOps to solve business problems at scale. The Seldon team is built from world-class talent who work closely to understand the theoretical and technical challenges and the business and operational challenges that customers face daily. Through this funding, coupled with an immense amount of passion and grit, we’ve subsequently built a suite of tools that help organisations overcome these challenges. In addition, we have expanded our leadership team over the last year to include Tanu Chellam (VP of Product Management), and most recently, James Perry as CFO at Seldon. He will steer our trajectory for continued success.
- We established our position as an industry leader in open-source MLOps without a dedicated community manager. Recently, our Head of Developer Relations Ed Shee joined the team to support the growing community and further engage and educate our users, contributors and customers. This community is the foundation of many of our industry-leading projects and has allowed the Seldon toolkit to constantly evolve and improve alongside the changing demands of our users.
- Over time, Seldon has created solutions that integrate seamlessly with other leading ecosystems, platforms and technologies such as Kubeflow, Red Hat OpenShift and HPE GreenLake. We’re dedicated to customer success. Knowing that Seldon plays a key role in many organisations’ wider tech stacks, we’ve worked tirelessly with our partners and customers to refine and extend our offering.
Shaping the Future
- We’ve proven our value to the market. With our early adopter customers, Seldon Deploy has reduced project lifecycles from months to days – or even hours once the system is embedded as part of an organisation’s workflow. This enterprise product streamlines production workflows, including risk management through fully integrated model monitoring, explanation and auditing workflows. One organisation in the Fortune 100 reported that Seldon’s technology helped save tens of millions of dollars on its machine learning projects in cost savings alone, not taking into consideration the enormous value that the models bring to their business. Seldon Deploy 1.0 launched in Feb 2021 with guest appearances from clients AstraZeneca, H&M and Red Hat, who shared the key part that MLOps plays in accelerating their business.
- We’re particularly proud of the part that Seldon has played in the last 18 months in accelerating drug discovery and medical imaging within some of the world’s leading life sciences organisations.
- Seldon has proven productivity gains of up to 92% for organisations deploying machine learning models. Our technology helps enterprise customers take their machine learning projects from POC to production with maximum efficiency and minimal risk. Just ask some of our leading customers, including Johnson & Johnson, H&M, and Covea Insurance.
We’ve recruited a world-class team of over 40 people – and now we’re growing it further. We even hired our first-ever Head of People and Talent to make it happen. This strategic appointment is driving forward our ambitious growth plans for Seldon while helping us to maintain and strengthen our first-class company culture. As a result, Seldon’s headcount will double again over the next year.
Interested in joining Seldon and becoming part of our next seven years? Visit our careers page.