Financial services and insurance

Where we've found AI belongs

Each of these comes from work we've done in the sector, not from a list of possibilities.

• Sales teams that learn from every conversation

  Feedback that reaches each salesperson from their own calls, while it can still change the next one.

  Caliper→
• Research that finds its own opportunities

  A research team maps how its work actually runs and sees where AI would help, ranked by impact and feasibility.

  Compass→
• Customer conversations

  A voice agent designed around what customers call about, and what a person should still handle.

  Workbench→
• The work around the work

  Internal tools that take on the planning and busywork, like preparing the events a sales team hosts.

  Workbench→

• We brought AI into sales coaching, event planning and customer service.

  We built AI-powered feedback for their sales team, internal tooling that plans the menus for the restaurant seminars they host, and designed the future of their customer-facing AI voice agent. One new AI product went from concept to production in under a month.

  

• We helped a market research team find where AI belongs in its work, then start building it.

  We worked through how their research actually gets made and mapped it with them. Then we helped the team start prototyping against what the map turned up, both in our tools and in the ones they already had inside the firm.

  A global investment firm

What your team gets

• Your experts decide what good looks like

  The people who do the work write the standard an answer is judged against, and they can do it from a link without an account.

• Answers from your own material

  Policies, procedures and documents you control, so people get your firm's answer rather than the internet's.

• You can open the run behind an answer

  Every run is kept with what it did and what it cost, so nobody has to take the result on faith.

• You can show what it changed

  Write down what you expect a change to move before it ships, then settle it against the numbers your systems already produce.

What your risk and compliance teams will ask

The people responsible for your systems decide what AI may touch, and nothing changes without a person agreeing to it.

• Runs on your own infrastructure

  The Workbench SDK runs flows entirely inside your own systems, with your own model keys.

• Choose the models

  An Enterprise workspace can limit which model providers its flows are allowed to use.

• Nothing changes without approval

  zv1 shows the exact edit and waits for a person to approve it. Admins decide how freely it may use each connected tool.

• Keys stay in the vault

  Keys for the services your flows call live in the workspace vault, not in the flows that use them.

• Every version is kept

  Workbench keeps a version history for every flow, and your code can pin the exact version it runs.

• Single sign-on

  On the Enterprise plan, your team signs in with the identity provider your company already uses.

Three ways of working with us:

• Inside your team

  We join your team to find the work worth changing, build it with you, and prove it holds before it scales.

• Alongside you

  We work with your team a couple of times a week, hands-on with the technical problems and the strategic ones.

• With our tools

  Your whole team works in the tools we use on our own client work, in one shared workspace.