Autype: create & automate documents.Try it
Back to blog
Trades and construction craft06/06/2026

AI Automation for Trades: How a Business Saves 18 Hours per Week with HITL Agents

A 28-employee trades company automated its quote creation process with centerbit AI agents. The process automation results: 18 hours saved per week, zero calculation errors, and 37 percent more quotes. The full AI automation case study with metrics, testimonials, and success factors for trades businesses.

The Starting Point: A Trades Business at Its Limit

One of our customers, a mid-sized trades company with 28 employees in southern Germany, specializes in drywall construction, interior finishing, and energy-efficient renovation. The company handles around 50 customer inquiries per month, produces roughly 35 quotes, and simultaneously manages 20 to 25 active projects. A typical scenario for the construction industry: high order volume, tight staffing, and growing cost pressure.

The problem: quote creation and order processing automation was virtually nonexistent. Customer inquiries for quotes arrived via email, phone, or the website contact form. A senior project manager read each inquiry, researched material prices, calculated labor hours, wrote the quote, and entered everything into the ERP system. Each quote consumed an average of 45 minutes: a massive time drain that tied up 26 hours every week. Resources urgently needed elsewhere.

Baseline
26 h
weekly time spent on quote creation
Error Rate
12%
quotes with calculation errors or outdated pricing
Turnaround
2.8 d
from inquiry to quote delivery

Compounding the problem: material prices in the trades changed weekly, yet the project manager often calculated with outdated lists. The result was calculation errors in roughly 12 percent of quotes, leading to either renegotiations or margin erosion. Meanwhile, little time remained for construction site supervision — the project manager was chained to the office. A classic digitization problem familiar to many trades businesses.

The Solution: Process Automation with HITL Agents

Instead of a fully autonomous AI system, the company chose a Human-in-the-Loop approach with centerbit. The AI agent handles the structured groundwork for quote creation; final approval and quality control stays with the human. A proven model for AI automation in SMEs. The solution was rolled out in three progressive phases:

1
Phase 1: Automated Inquiry Capture and Classification
The AI agent reads incoming emails and contact form submissions automatically. It extracts trade type, project scope, address, and urgency, and classifies the inquiry into one of four categories: standard quote, rough estimate, complex project, or spam. The classification is presented to the project manager as a suggestion — they can confirm or adjust with a single click.
2
Phase 2: Automated Calculation with Live Material Prices
The agent accesses the building supplier's current price list and calculates material and labor costs based on project data. It creates a complete quote draft and submits it with full calculation details for review. Three critical checkpoints prevent faulty quotes from slipping through: quantity plausibility check, current supplier price verification, and formal completeness validation.
3
Phase 3: Approval and Dispatch
The project manager reviews the draft in the centerbit interface, can adjust prices, add line items, or leave comments, and releases the final quote with a single click. The agent sends the PDF, logs it in the ERP system, and schedules an automatic follow-up reminder after seven days for quote tracking.

The Results: 18 Hours Saved per Week

After three months in production, the process automation with AI agents delivers measurable results across the entire quoting pipeline:

MetricBeforeAfterChange
Weekly time on quote creation26 hours8 hours-69%
Turnaround: inquiry to quote2.8 days6 hours-93%
Calculation errors12%<1%eliminated
Quotes per month3548+37%
PM presence on construction sites4 h/day7 h/day+75%
From the Project
"At first, I was skeptical whether an AI could truly understand our quotes. Today, I wouldn't go back. The agent handles all the grunt work — I just review and approve. Trust built over time: after three weeks, I knew exactly which areas needed closer attention. Now I'm on-site more than in the office, with 20 percent more orders coming in."
Project Manager at the company

The Success Factors: Why HITL Makes the Difference

Three factors were decisive for the project's success in AI automation for trades businesses:

1. Phased rollout, not big bang. Phase one ran for four weeks as an isolated classification aid. Only when the project manager trusted over 95 percent of the classifications did Phase 2 follow. This staged implementation built trust and acceptance rather than provoking resistance — a key success pattern for digitization in the trades.

2. Configurable approval thresholds. Not every quote goes through the same process. Standard quotes under €5,000 in volume pass through with simplified review; complex projects over €20,000 require detailed manual verification of every line item. This risk-based tiering saves time on quote creation without compromising quality.

3. The agent learns from every correction. When the project manager repeatedly adjusts a specific value, the agent adopts the pattern. After three months, the rate of manual corrections dropped from an initial 35 percent to under 8 percent. The AI becomes more precise with every use, increasingly freeing people from routine tasks.

What Other Trades Businesses Can Learn

This customer's experience is no isolated case. The underlying process automation patterns — extracting structured data from unstructured inquiries, rules-based calculation with live data, HITL approval at critical points — are transferable across numerous trades: from electrical contractors to HVAC installers to roofing companies.

What matters for successful AI automation in the trades isn't company size but the willingness to standardize processes. Where workflows are already documented and structured, a HITL agent can go live within two to three weeks. Where processes exist informally in the owner's head, that groundwork comes first.

Getting started with AI-powered quote creation is significantly easier today than a year ago. Standardized agent templates and preconfigured industry workflows reduce setup time from weeks to days. And the Human-in-the-Loop approach ensures that no automated quote ever leaves the building without human eyes on it.

The math for process automation in the trades is straightforward: 18 hours per week, 52 weeks per year, at an hourly rate of €85. That's nearly €80,000 flowing from the office back to the construction site — and into more orders, better quality, and happier customers.

centerbit

Book a consultation now

If you see similar manual work in your team, we can review the process together in a free initial consultation.

Request consultation
AI Automation for Trades: Save 18h/Week with HITL Agents | centerbit Case Study | centerbit