Contract review experts Robin AI have announced a deep partnership with Amazon Web Services (AWS) that will see it engage with the platform’s Bedrock and SageMaker training and experimentation capabilities, as well as joining the AWS Marketplace, where customers can test and deploy their genAI offering.
In January the company completed a $26m Series B funding and now has over 150 staff. It has also just opened an office in Singapore to serve the Asia-Pacific market. The AWS deal marks another step in the business’s expansion drive.
In terms of working with AWS, Robin AI, which has a strong relationship with Anthropic and has pioneered the use of their Claude LLMs for contract analysis, will now use Amazon Bedrock ‘to build and scale generative AI applications using [the] Claude models while keeping customer data completely private, a critical step to help legal teams overcome their hesitancy to adopt AI’.
But, what is Bedrock? Here is what Amazon says: ‘Using Bedrock, you can easily experiment with and evaluate top Foundational Models (FMs) for your use case, privately customize them with your data using techniques such as fine-tuning and Retrieval Augmented Generation (RAG), and build agents that execute tasks using your enterprise systems and data sources. Since Amazon Bedrock is serverless, you don’t have to manage any infrastructure, and you can securely integrate and deploy generative AI capabilities into your applications using the AWS services you are already familiar with.
‘The single-API access of Amazon Bedrock, regardless of the models you choose, gives you the flexibility to use different FMs and upgrade to the latest model versions with minimal code changes.’
In short, on top of experimenting with legal use cases more easily you can really focus on boosting and measuring accuracy – and right now, getting better at accuracy is vital to being a successful legal genAI provider.
Meanwhile, the UK-based company, which is also expanding into the US, is using Amazon SageMaker to ‘accelerate the building, training, and deployment of its machine learning models’.
In terms of SageMaker the tech giant says: ‘…you can build, train and deploy ML models at scale … all in one integrated development environment (IDE). In addition, you can build your own FMs, large models that were trained on massive datasets, with purpose-built tools to fine-tune, experiment, retrain, and deploy FMs. SageMaker offers access to hundreds of pretrained models, including publicly available FMs, that you can deploy with just a few clicks.’
And, using AWS’s ability to store data in specific regions, Robin AI stated that it ‘can quickly create custom versions of its products for different countries, including the US, UK and across the EU [while it] guarantees that confidential customer information never leaves AWS’s secure cloud environment and is stored in compliance with local data protection regulations’.
All well and good, but what does this mean practically for Robin AI and its clients?
James Clough, co-founder and Chief Technology Officer of Robin AI, explained: ‘Thanks to AWS, we don’t have to compromise between powerful and safe AI – our customers can have both. Amazon Bedrock gives Robin the ability to make a rock-solid global promise to our customers: with Robin you get access to the best AI models, such as Anthropic Claude Sonnet 3.5, while knowing your data is safe.
‘[And] Amazon SageMaker allows us to build generative AI capability directly into our product, speeding up the development of new features, and setting us apart from legacy companies in legal technology markets.’
And this may well be key as LLMs continue to advance at pace, leaving what older models could do in the rear view mirror and demanding legal tech companies match what can now be achieved to their product.
The legal tech company also stressed that its approach ‘allows customers to go further than if they were using only a Large Language Model (LLM) alone’ because they are also wielding their own ‘proprietary AI models that are trained on over 100 million contract clauses, data which is typically not publicly available for LLMs to train on’.
And this additional bit is really where all the experimental capability provided by AWS comes into play.
In short, with this extra legally-focused capability, plus the adaptability for implementing new advances in LLMs via the AWS offerings above, and plus the ability to ringfence data processing to keep clients happy, the company is hoping to further expand its business in the increasingly competitive field of legal genAI. After all, contract review is now one of the most frequently cited ‘skills’ offered by a range of legal tech companies that are tapping genAI.
AWS is also rather pleased to have a growing legal tech company as a customer. Baskar Sridharan, Vice President of AI and Infrastructure at AWS, added: ‘We are inspired by how companies like Robin AI are incorporating generative AI to drive transformation within the contract review process so legal teams can focus on their priority work. It’s exciting to see how Robin AI is leveraging Amazon SageMaker and Amazon Bedrock to build, train, deploy and scale their gen AI-powered legal assistant to further their mission of supercharging the way legal work gets done.’