Harvey, a generative-AI assistant for First Technology’s legal team
A cloud chat assistant, Harvey, that lets legal and paralegal staff check contract terms, dates and notice periods in seconds instead of digging through documents.
Legal and paralegal consultants at First Technology head office spent a lot of time querying supplier, client and landlord contracts to check terms, termination dates and conditions. It was a slow part of the legal cycle, and it often relied on senior consultants, which held up the work.
02
The approach
We found a clear business case for giving the legal team a generative-AI assistant, and deployed Harvey within a week using our template-based chatbot framework on an Azure AI RAG architecture. It runs on the First Digital AI RAG framework over Azure AI Services, so contract data stays inside First Technology’s own Azure tenant, with Entra ID limiting access to authorised users. As with our other AI work, we kept it secure and cost-aware with an AI-workload landing configuration and the most cost-effective model from the Azure AI Foundry catalogue.
03
The outcome
The in-house legal team now has fast, intelligent access to the knowledge already sitting in its contracts. Queries that used to take a senior consultant’s time get answered in seconds, and that knowledge is captured for the future. It also opens the door to contract comparison, pre-screening supplier contracts and guidance on NDAs.
Summary & benefits
Fast answers on contract terms, dates and notice periods.
New legal assistants lean less on senior colleagues for contract knowledge.
Sets up contract comparison and supplier pre-screening next.
Similar work in Technology & Innovation
More customer stories like this one.
Private token, no per-seat cost, company wide impact!
We gave First Technology staff local AI chat and agentic desktop agents running on our private NVIDIA GPU cluster, so the data and token generation stay local, the bill is in local currency, and responses come back fast.
Vector Gaming needed serious GPU horsepower to train a custom 350-million-parameter model for smarter game NPCs. We put our private GPU cluster behind their Vulkan-based distributed trainer and cut a two-month run down to a fraction of that.
Urban Connect’s agentic diagnostic agent works a network fault out for itself, pulling the logs, querying the routers and upstream providers, then writing the call-centre ticket, so support is proactive instead of a manual scramble.