DM Response Agent
Headache. Leads sit in Instagram and Facebook DMs for hours.
Answers common questions in your voice, qualifies prospects, and routes hot leads to a human or booking flow.
Practical AI · Real business outcomes
Practical AI for the small business headaches that cost you leads.
Token Train builds focused agents that answer DMs, recover missed calls, follow up with leads, and chase reviews — so you stop losing revenue to silence.
🚂
tokentrain
Typically replies instantly
The headaches we hear every week
- Missed Instagram and Facebook DMs
- Slow response to new leads
- Forgotten follow-ups
- Missed calls that turn into lost revenue
- Reviews left unanswered
- Quotes sent but never followed up
What we build
A dozen agents for the work that eats your day.
Each one owns a single, named headache — a clear trigger, a defined action, and a number you can watch move. Six are live in the playground right now.
Capture every lead
The moment someone reaches out — a DM, a missed call, an email — an agent is already answering.
Missed Call Recovery
Headache. Every missed call is potentially a lost customer.
Sends an immediate text after a missed call, captures what the customer needs, and creates a follow-up task.
Shared Inbox Triage
Headache. Important emails get buried and the owner becomes the bottleneck.
Sorts incoming email, summarizes threads, drafts replies, and routes messages to the right person.
Fill the calendar
Turn interest into booked appointments — and make sure the chairs don't sit empty.
Appointment Booking Agent
Headache. Booking is a game of phone tag and DMs at midnight.
Offers your real open slots, answers questions, confirms the appointment, and writes it straight to your calendar.
No-Show Reduction
Headache. Empty slots and last-minute cancellations quietly cost revenue every week.
Sends confirmations and reminders, offers one-tap reschedules, and flags likely no-shows before they happen.
Measured by
- · No-show rate
- · Reminders confirmed
Waitlist & Cancellation Filler
Headache. A cancellation is an empty chair unless someone fills it fast.
The moment a slot opens, the agent offers it down your waitlist in order and books the first yes.
Measured by
- · Slots refilled
- · Revenue recovered
Get paid faster
Follow up on the quotes and invoices that stall out — without the awkward chasing.
Lead Follow-Up Agent
Headache. Warm leads go cold while you're busy running the business.
Sends the first reply within minutes, personalizes follow-up, and stops the sequence the moment a lead responds.
Quote & Estimate Follow-Up
Headache. Most quotes never get a second look after you hit send.
Follows up on unaccepted estimates, answers objections, and flags the deals worth a personal call.
Measured by
- · Quote acceptance rate
- · Follow-ups sent
Invoice & Payment Reminder
Headache. Chasing overdue invoices is awkward and easy to put off.
Sends polite, escalating reminders until an invoice is paid — and stops the second it clears.
Measured by
- · Days-to-payment
- · Overdue invoices cleared
Keep customers coming back
Protect the reputation you've built and pull repeat business out of customers who'd otherwise drift.
Review Response Agent
Headache. Negative reviews sit unanswered for days. Positive ones too.
Drafts on-brand replies, flags severe complaints for human approval, and surfaces repeated issues you should fix.
Referral & Testimonial Request
Headache. Happy customers will refer you — if you ask at the right moment.
Spots the moment a customer is delighted, asks for a review or intro, and routes warm referrals to you.
Measured by
- · Reviews requested
- · Referrals captured
Win-Back & Reactivation
Headache. Customers drift away quietly and you never notice until it's late.
Watches for who's gone dark, reaches out with a real reason to return, and books them back in.
Measured by
- · Dormant customers re-engaged
- · Repeat bookings
What it's costing you
$36,000 a year is a typical small-business cost of silence.
Unanswered DMs, missed calls, late email replies — every one is a quiet leak in your pipeline. Run the numbers for your business in 30 seconds.
Sample: a 300-inquiry / mo business
$3,000/ month
Today$3,000
With Token Train$375
That's $31,500 recoverable per year.
How it works
Every Token Train agent is designed around five elements.
Not a chatbot. A working system with rules, escalation paths, and measurable outputs.
- 01
Trigger
A specific event starts the workflow — a new DM, a missed call, a form submission, an overdue invoice.
- 02
Decision
The agent reads intent, urgency, and risk before acting. It knows the difference between a sales lead and a complaint.
- 03
Action
Reply, route, tag, book, summarize, or update CRM — whatever the workflow actually requires.
- 04
Guardrails
Never invents pricing, policy, or availability. Stays in your brand voice. Refuses anything risky.
- 05
Escalation
Hands off to a human the moment a customer is angry, a deal gets serious, or confidence drops.
Why Token Train
Built for real workflows, not demos.
01
Narrow and useful
One headache per agent. No 'AI transformation' projects — just focused systems that remove daily friction.
02
Measurable from day one
Response time, qualified leads, no-shows recovered, hours saved. If we can't measure the lift, we don't ship it.
03
Escalates intelligently
Agents know when to step aside. Angry customers, VIP clients, and edge cases reach a human fast — with full context.
FAQ
What people ask before booking.
How long does it take to ship an agent?
Most first agents are live within 2–3 days, then fine-tuned and fully operational within 2–3 weeks. We scope narrow on purpose — one workflow, one outcome — so we can deploy fast and measure the lift before expanding.
What systems can you connect to?
Almost 1,000 SaaS tools — including Instagram, Facebook Pages and Messenger, Gmail, Outlook, Google Calendar, Calendly, Cal.com, Slack, HubSpot, Salesforce, Pipedrive, Notion, Google Sheets, Airtable, Stripe, and Zendesk. Our first live workflow is on Instagram. One caveat: Meta-owned platforms (Instagram, Facebook Messenger) require Meta App Review and business verification before agents can send messages at scale — we handle the paperwork, but plan for a few weeks of platform approval on top of build time.
What happens if the AI gets something wrong?
Every agent has explicit guardrails and an escalation path. Risky messages, angry customers, refund demands, and uncertain answers go to a human with full thread context — never auto-resolved.
How is this priced?
A one-time setup to scope, build, and deploy the agent, then a flat monthly fee covering hosting, monitoring, tuning, and support. We share concrete numbers on the discovery call once we understand the workflow.
Will replies actually sound like us?
Yes. We train every customer-facing agent on real replies from you and your team. We test for AI tells — generic openers, robotic phrasing, over-enthusiasm — and tune them out before going live.
Get started
Stop losing leads to silence.
20 minutes. We'll map the one or two workflows where AI will pay for itself first — no slides, no pitch.
Prefer to write first? Use the form — we read every one.