It's 9:47 PM. A homeowner texts about a deck project. By the time you see it, she's already booked someone else. What if someone answered for you?
10-minute setup. No app to download. Just text.
of customers choose the contractor
who responds first
more likely to qualify a lead
with a 5-minute response
average contractor response time
DeskForeman responds in seconds
Maya is a DeskForeman AI agent. She talks to your customers, qualifies leads, and keeps you in control.
What your customer sees
What you see
Maya, a DeskForeman AI agent, handles it all. You just build.
Replies in seconds, 24/7. No lead left waiting while you're on a job site.
Scores every lead 0-100 automatically. No more wasting afternoons on tire-kickers.
AI-generated pricing from project details. Share with customers or review first.
Coordinates and books visits straight to your Google Calendar.
Drafts full proposals and contracts from site visit data. You approve, she sends.
Daily insights: stale leads, conversion patterns, and what to focus on next.
Maya starts by running everything by you. As you approve her work, she earns more independence — just like managing a real employee.
Every estimate and action goes through you first.
Routine tasks handled. Unusual items flagged for review.
Maya runs your pipeline. You get alerts only on exceptions.
"You've approved 8 out of 9 estimates I flagged. Want me to start sharing ballpark ranges on my own? I'd still flag anything unusual."
One recovered lead pays for years of DeskForeman.
No tiers. No per-message fees. No contracts. Cancel anytime.
| DeskForeman | Chatbots | Contractor CRMs | Virtual Receptionist | |
|---|---|---|---|---|
| Price | $499/mo | $500-2,000/mo | $50-200/mo | $300-600/mo |
| Interface | SMS | Web widget | Dashboard | Phone |
| Lead qualification | Automatic, scored | Rules-based | Manual | None |
| Estimates | AI-generated | None | Manual entry | None |
| Proposals & contracts | AI-generated | None | Templates | None |
| Setup time | 10 minutes | Days-weeks | Hours-days | Hours |
| Builder adoption | Texts from day one | Requires training | 30-40% | Works, no intelligence |
No app. No dashboard. No training. Just text.
DeskForeman isn't a single chatbot. It's a coordinated team of specialized agents, each running the best-fit model for its job.
| Agent | Role | Optimized For |
|---|---|---|
| Customer Assistant | Talks to customers, extracts lead data | Quality + natural tone |
| Lead Qualifier | Scores leads 0-100, routes decisions | Speed + accuracy |
| Initial Estimator | Generates ballpark pricing | Reasoning |
| Response Reviewer | QA gate before every customer message | Speed (~200ms) |
| Escalation Detector | Flags frustration, alerts builder | Signal detection |
| Visit Scheduler | Books site visits via Google Calendar | Coordination |
| Proposal Generator | Drafts full proposals from visit data | Document quality |
| Contract Builder | Generates contracts from approved proposals | Precision |
| Pipeline Coach | Daily insights and coaching | Analysis |
| Onboarding Agent | Builder setup and management via SMS | Conversational memory |
High-reasoning tasks (proposals, customer conversations) run on frontier models. Classification tasks (lead scoring, QA) run on speed-optimized models at 10-15x lower cost per call.
The customer assistant returns conversational response and structured lead data in one API call — cutting latency and cost in half versus a two-call pattern.
Background automation only fires when new information is shared. Chitchat doesn't trigger the qualifier or estimator — preventing wasted compute.
Circuit breakers on every dependency. Exponential backoff retries. Graceful degradation — if the AI API goes down, customers get a friendly hold message, not silence.