Introducing AWSSfn: CaseMark's Latest Contribution to the Open Source AI Community

CaseMark is committed to contributing to open source AI

 min. read
July 2, 2023
Introducing AWSSfn: CaseMark's Latest Contribution to the Open Source AI Community

Pioneering the Path with LangChain and AWS

At CaseMark, we are on a mission to develop advanced AI solutions for legal professionals, but want to do while also helping push the AI industry forward. As we continue on our journey in developing a robust AI platform for the legal services industry, we’ve committed to contributing to the open source ecosystem along the way. Our first of many contributions is a tool that expands the capabilities of popular AI developer framework – LangChain. We're proud to announce the release of AWSSfn, a tool that allows application developers to integrate AWS Step Functions within their Large Language Model (LLM) applications. This will allow application developers to automate complex business workflows, build domain-specific AI applications, and build countless other software solutions powered by AI.

Empowering Enterprise Use-Cases

Larger enterprises usually have to take a more reserved approach when it comes to the adoption of new technologies. These organizations have to weigh different factors and concerns like compliance, governance, security, etc. before jumping in to the latest bleeding edge tech. However, one way to reduce this friction is to provide simple onramps to these new technologies. The introduction of AWSSfn is just one example of an onramp that will allow enterprises already leveraging Amazon Web Services (AWS) to integrate generative AI into their technology stack. LLMs have just as many potential use-cases for enterprises as they do for end-users.

Finite State Machines and LLMs: Exploring Uncharted Territories

AI is a relatively young industry with many unknowns. As individuals continue to push the boundaries of what is possible with this technology, we have to stay cognizant of challenges faced throughout the adoption lifecycle of previous technologies. An inherent characteristic of  complex LLM applications is that they exhibit unbounded behaviors. The proliferation of generative AI will undoubtedly introduce new challenges. We’ve started to think deeply about many of the challenges that have already surfaced. Our current research lies at the intersection of Finite State Machines (FSMs) and LLMs to introduce state boundedness and controlled computational complexity within generative AI applications. We’ll be using AWSSfn as part of our ongoing research into this problem space. Stay tuned for our published research paper expected to be released to the general public in the coming months.

What's Next?

We will continue to make contributions to LangChain on multiple fronts starting with the development of a Python version of the AWSSfn tool to reach even more developers. We are actively seeking feedback and feature requests from the community. If you have any ideas, comments, or feedback please reach out to us. We’d love to build alongside you 😃.

Join Us onThis Exciting Journey

This open-source tool is just one of our contributions to the growing ecosystem of generative-AI open source software. As we continue to lean heavily into LangChain, we invite you to join us in this journey. Follow our updates on Twitter (@casemarkai), and let's explore the possibilities of AI, together.

At CaseMark, we're not just creating AI tools for the legal profession; we're hoping to help build the future of AI with our open source contributions.