Jylo, which provides an ‘internal AI Product Marketplace’ among other core features, is partnering with health sector law firm Hempsons.
Jylo, which was previously spun out of Travers Smith, said that its approach to the use of genAI tools has the ‘object of transforming existing intellectual capital into repeatable, AI-driven processes in a secure and confidential manner, and is customised for each practice area’.
Jylo’s commitment to data security and privacy was a decisive factor for Hempsons, they added. Through Jylo’s partnership with Microsoft for UK-based inferencing for AI, ‘the solution met Hempsons’ stringent requirements for a single-tenant model, ensuring end-to-end data sovereignty within the UK – a critical necessity for Hempsons’ clients’.
Data is safeguarded at all times and is not used for any product development or training work, remaining within Hempsons’ control at all times, they noted.
In addition, Hempsons’ practice groups can now create bespoke playbooks to ‘upskill users, standardising prompting for repeatable tasks and harnessing sector-specific expertise’. And Jylo’s workflow integrates guardrails, ‘maintaining lawyers as key contributors through a human-in-the-loop approval process’. This includes citations linking directly to document sources and a thorough record of actions, replicating traditional working methods with AI support where needed, they concluded.
And here’s an overview of what Jylo can do now:
- ‘Product Marketplace – Turn your organisation’s insight into an internal marketplace. Capture, retain and reuse institutional knowledge. Operationalise and democratise the expertise of your best people.
- Model Agnostic – Remain up to date and in control with model selection. Utilise the right model for the right job, managing costs whilst enabling scale.
- Compliant Chat – Block all consumer AI tools within your organisation. Offer Jylo as an alternative. Democratise model selection allowing employees to retain access to their favourite models whilst adhering to your security and privacy policies.
- Project Centric – Move away from AI as a copilot or chatbot for the individual. Embed as a team mate for your project. Project Centric Jylo supports team collaboration and more closely aligns to existing security within organisations.
- Dynamic Filters – Convert unstructured data into structured data. Map the structured data to bespoke custom filters for ease of navigation.
- Verification Workflow – With human-in-the-loop verification, eliminate the risk of hallucination. Don’t take AI’s word for it, quickly check, verify, amend or accept AI’s output.
- Text Analysis – Quickly copy and paste source of truth information from external sources to interrogate and analyse. Convert external text into documents and add to your project with one click.
- Prompt Development Studio – Ensure your prompt engineers have the right tools to do their job. Build complex prompts in our proprietary Jylo development studio (IDE). Go granular and select different models for different prompts, all from within the same product.
- Analytics – Have complete auditability and transparency of AI use within your organisation. Manage return on investment by reporting on usage and cost.’
Shawn Curran, CEO of Jylo, said: ‘The team at Hempsons have really embraced how they can leverage Gen AI to reimagine and deliver efficient and effective legal services. Hempsons’ forward-thinking approach is set to improve processes, simplify and automate manual tasks, and ultimately bring new value-added services to clients.’
While Adam Johnson, Head of Technology and Architecture at Hempsons, commented: ‘Our deep understanding of our clients’ businesses drives us to innovate the solutions we offer. Gen AI, and specifically the Jylo model, empowers us to enhance our capabilities and to optimise internal business processes using Jylo to drive efficiencies.
‘The Jylo model allows us to create standardised repeatable processes, which can be benchmarked against different AI models to measure accuracy and effectiveness before being put into operation. We are currently developing several use cases which are already starting to yield positive outcomes.’
You can find out more about Jylo here.