- hello@1ree.com
- London · Dubai
By department and industry
Each of these starts as a pilot: four to six weeks, fixed price, one agreed metric. None of them goes live without an evaluation set and a human review queue. Showing 12 agents.
Filter by department
Filter by industry
Reads the enquiry, enriches the company, scores it against your fit rules, drafts a reply and offers call slots. Low-scoring leads are routed, not binned.
Builds a first draft from your rate card, scope library and past proposals. A human sets the price and signs it off — the agent never quotes.
Answers from your documentation and order data. When it cannot, it escalates with a written summary so the human does not start from nothing.
Handles “where is my order” and returns eligibility end to end, with a hard limit on what it can approve without a person.
Invoices, purchase orders, contracts and compliance forms: extracted, validated against your rules, posted to accounting or the ERP, flagged when unsure.
Watches purchase orders and stock levels, chases suppliers on a schedule, and summarises what is late and what it will affect.
Drafts, adapts and localises content into English and Arabic, then stages it for approval. It never publishes on its own.
Reads GA4, ad platforms and the CRM, then writes the monthly narrative: what moved, what it cost, and what it means. A human edits before it is sent.
Answers questions over policies, SOPs, project history and Slack or Teams, with your permissions mirrored so nobody sees what they should not.
Handles viewing requests, keeps listings consistent across portals, and answers tenant enquiries out of hours with a defined escalation path.
Runs structured intake, checks for conflicts against your rules, and produces a matter summary for the fee earner to review.
Takes reservations, answers venue questions, and handles amendments within limits you set, in English and Arabic.
No agents match that combination yet — try a different filter, or tell us what you need built.
Before any of them ship
| Ships with | What it means |
|---|---|
| Evaluation set | 50–200 real examples from your own records, with what a good response looks like. Yours, in your repository. |
| Guardrails | Prohibited actions, spend limits and escalation triggers, agreed before the agent runs. |
| Human review queue | The agent drafts. A person approves. Autonomy expands only as evaluation scores earn it. |
| Logging | Every output traceable to its inputs, including which model version produced it. |
| Named models and regions | Written into the contract, so compliance can answer the question without calling us. |
| Handover option | Training so your team can maintain the simpler agents without us. |
Next step
That is what the AI Opportunity Audit answers. Two to three weeks, fixed price, and you keep the document whatever you decide to do next.