Scoring · routing
Lead Qualification
Every inbound enquiry is scored with evidence, checked by deterministic rules and routed: full offer, starter, needs info, review or reject.
The problem
Enquiries arrive from forms, email and calls, mixed with spam and tyre-kickers. Answering all of them by hand is slow, and an LLM that "just decides" can't be audited and can be talked into anything.
What we build
- Leads from your contact forms, forwarded email, website chat, Telegram, calls or an n8n webhook, all scored the same way.
- An LLM that rates fit, budget clarity, urgency, authority and scope, quoting the lead's own words as evidence.
- Code that extracts budget, currency and timeline ("ASAP", "next quarter", "before Christmas") instead of trusting the model.
- A rules gate that decides the route, a reply drafted from your templates, and alerts (Telegram, Slack, email) for the leads worth your time.
- One inbox to reply from, plus a signed webhook into your CRM or a Google Sheet.
Guardrails
- The model scores; code decides. Even a model fooled into giving a score of 100 can't pass the gate.
- Spam and injection are detected deterministically and never get a sales reply.
- Replies never invent prices, dates, availability or promises.
- Every decision is written to an audit log with reason codes.
See them run on real data.
The apps run real models on every click, so they sit behind a password. Ask for a login or a guided walkthrough: talk to our AI receptionist in Front Desk, drop invoices into Docket, or load forged applicants into Dossier. Or try them in public: score a lead, read your invoice, forge an applicant, or connect Docket and Dossier to Claude. Or hire us to build yours.