Tessoryn AI

AI agents for customer operations

AI agents for customer operations.

Tessoryn builds AI agents that resolve customer requests across the tools your teams already use.

Agents can understand a request, retrieve context, use authorised tools, complete an action and report exactly what happened.

Product UI prototype · replace with current Tessoryn build
Refund operations
Ready
Customer request

“Resolve this customer’s refund request.”

Order #18427 · received just now
  1. Received requestTask created and customer identified.
    Queued
  2. Retrieved customer recordAccount context loaded from permitted customer data.
    Waiting
  3. Checked orderOrder and payment status inspected.
    Waiting
  4. Applied refund policyRequest evaluated against workflow rules.
    Waiting
  5. Issued refundAuthorised action executed through the connected payment tool.
    Waiting
  6. Updated support ticketOutcome recorded for the support team.
    Waiting
  7. CompletedCustomer notified and action summary created.
    Waiting
Task → status → tool usage → action → result

Product

Operational software, not another chatbot.

Tessoryn is designed around work that crosses systems. The agent receives a task, gathers the context it is allowed to use, takes authorised actions and leaves a clear record of the outcome.

01

Task

A request enters from a customer or operator with a clear outcome to achieve.

02

Context

The agent reads only the customer data, policies and workflow context required for the task.

03

Action

Connected tools are used within explicit permissions. Sensitive steps can stop for human approval.

04

Result

The agent reports what it changed and leaves an execution history for review.

Product walkthrough

See one workflow from request to completed action.

This slot is intentionally labelled as a prototype until an approved Tessoryn product recording is supplied. Replace the poster and video source with a 30–60 second capture of the current product.

Recording not published yetReplace /public/product-demo.mp4 when an approved recording is available.

How it works

From request to resolved operation.

Each workflow has a defined objective, known systems and explicit action boundaries.

  1. 01

    Receive the task

    Start from a customer request, queue item or operator instruction.

  2. 02

    Understand context

    Identify intent and retrieve the information required to make the next decision.

  3. 03

    Use connected tools

    Read or update business systems through the permissions assigned to the workflow.

  4. 04

    Complete and report

    Finish the authorised action, record the steps and return a concise outcome.

Security & control

Autonomy needs boundaries.

Tessoryn’s product model separates access, actions and approvals so operators can understand what an agent is allowed to do. We do not publish certification claims until they are independently verified.

Controlled system access

Scope the data and tools available to a workflow rather than granting blanket access.

Permission-aware actions

Keep actions within the authority defined for the workflow and connected system.

Human approval points

Place review gates before sensitive or exceptional actions where human judgment is required.

Action history

Preserve a reviewable sequence of task steps, tool calls and completed actions.

Read the security overview →

Company

Building Tessoryn in London.

Tessoryn AI is an early-stage company focused on agents for customer operations and multi-step business workflows.

London, United Kingdom contact@tessoryn.com X / @Tessoryn ↗
01

David Mabbott

Founder & CEO

Artificial Intelligence · University of Southern California
02

Pravesh Gpr

Co-Founder

Robotics AI & Autonomous Systems · University of London
Company & leadership →

Product demo

See Tessoryn resolve a real workflow.

Bring a customer operation your team currently handles by hand. We’ll use it to frame the demo around a concrete workflow.

Book a demo