AI quote calculation: Calculate faster in the trades, decide with confidence

A photo of the installation situation, a material list and an old order are enough – seconds later the AI proposes a finished quotation text including quantities and price. For an Austrian craft business that sounds like less paperwork in the evening. It becomes dangerous when a plausibly worded draft is mistaken for a reliable calculation. A forgotten setup allowance, an outdated purchase price or an invented line item can turn a quick quote into a loss-making job.

Properly used, AI in quotation calculation is not a price machine but a structuring assistant. It can organize approved information, formulate follow-up questions, compare line items and draft a clear accompanying text. Cost assumptions, scope of services, risk and approval remain with the people who know the company and the job. This guide shows how carpentry, installation, electrical, metal and other trade businesses can set up such a process practically, data-minimizing and traceably.

Why quotes are particularly error-prone

A quote links technical surveying, scheduling, materials management, labor and overhead costs, liability issues and customer communication. Much information is distributed: handwritten measurements, photos, supplier lists, previous calculations, emails and experiential knowledge. AI can bring this variety into a unified structure faster. But it does not know whether an old price is still valid, whether a wall is actually load-bearing, or whether the access route requires a second person.

The WKO notes in its calculation aids that businesses must work with their own concrete company data. That is exactly the guideline for AI: a model can recognize patterns, but it does not replace current cost bases or the operational decision. A quote must be based on verifiable inputs. If a basis is missing, the system should flag a gap – not invent a value.

A low price is not a sign of quality either. A sound calculation takes into account direct costs, productive working time, overhead, subcontracted services, contingency, profit and the specific utilization. The AI may explain this logic and prepare calculation steps. Which approach is economically correct and defensible to customers is decided by the business.

The safe role shift: from pricing decision to verification aid

The best task description is not "Calculate the correct price", but "Organize the approved data, perform the specified calculation procedure, and show every assumption." This shifts the role of the AI. It delivers a draft that experts can review in a targeted way.

Suitable tasks include:

Transfer measurement notes into a predefined item list;
Mark missing unit, quantity, performance limit, or delivery deadline;
Standardize material descriptions without replacing products;
Check a calculation against established plausibility rules;
List deviations from an approved comparison order;
compose a clear customer-facing text from already confirmed line items;
generate a list of outstanding technical questions for the final review.

Autonomous discounts, made-up quantities, automatic commitments to standards or deadlines, and price changes without a verifiable reason are not suitable. Especially for flat-rate prices, it must remain visible which services are included, excluded, or offered only as an option.

Which data need to be in order first

Poor master data are not improved by AI, only processed faster. Before a pilot, companies should therefore clarify five data areas: current purchase prices and their validity, their own hourly and machine cost rates, typical time values, overhead surcharges, and clear service components. Every number needs a source, a reference date, and a responsible role.

A usable dataset distinguishes at least between binding value, empirical value, and estimate. "Fitting: 84 euros, supplier list from September 8" is verifiable. "Fitting approximately 80" is an assumption. The AI must mark both types differently. Equally important are units: linear meters, square meters, pieces, and hours must not be silently converted into one another.

For personnel costs, current collective bargaining and company-specific bases are relevant. The WKO provides templates and calculation aids for various trades, but stresses that these are for orientation only. A model must therefore not adopt industry average values unexamined as the company's own truth. Accounting, payroll, and management provide the reliable figures.

A process in seven controlled steps

Define scope of work: Trade, location, requested service, interfaces, options and exclusions are clearly described.
Confirm intake: Measurements, photos and technical details are checked by a responsible person. Illegible or contradictory notes remain open.
Minimize data: Names, home addresses, faces, phone numbers and internal margins are only processed if they are truly necessary for the defined purpose and the processing workflow is authorized.
Create AI draft: The system only arranges the provided data, uses the specified calculation logic and lists assumptions separately.
Perform four checks:Technical review, quantity review, commercial review, and data protection review are assigned to appropriate roles.
Approve quote:An authorized person decides on price, risk, delivery date, and wording. The AI draft itself is never sent directly.
Post-calculation:After completion, target and actual values are compared. Insights are incorporated into time values and templates in a controlled manner, not automatically into a public model.

This process does not have to be cumbersome. For a small standard order, several checks can be performed by the same qualified person. The decisive factor is that the review takes place consciously and remains traceable later.

A prompt that leaves no gaps

“Create a quote structure from the following approved information. Use only the stated quantities, units, time values, and prices. Do not invent any items and do not change any values. Assign each number its source and status. List missing information, contradictions, and assumptions under ‘Review required’. Calculate subtotals according to the provided formula and show the calculation path. Do not provide any legal, technical, or commercial approval.”

After that, a second audit task helps: "Compare the draft and original data line by line. List every omission, reformulation with a potential change in meaning, and calculation discrepancy." This does not replace human oversight, but it makes it more targeted. Anyone who only asks for the final price easily overlooks an incorrect quantity estimate that coincidentally leads to a plausible sum.

Practical example: Built-in furniture for a Viennese office

A carpentry workshop is to provide a built-in solution with cabinets, a bench, and cable grommets. The master craftswoman records measurements, material preferences, surfaces, installation conditions, and two open points. The AI transfers this information into the company's item template and assigns materials, work steps, and installation time to each component. Where the wall condition has not yet been checked, it does not insert a made-up value, but rather a query.

When comparing with a previous year's order, the system recognizes that a similar bench was less deep. It does not suggest the same time value, but shows the deviation. Purchasing confirms current panel and hardware prices, the workshop management checks production hours, and the installation management takes access, floor level, and protective measures into account. Only then does management decide on overheads, risk, profit, and a possible discount.

In the customer text, the AI explains the confirmed services in an understandable way. The cable grommets are shown as an option; the unresolved wall reinforcement is not promised as a flat rate. This saves the company time in structuring and writing without handing over its experiential knowledge or pricing responsibility to a model.

Practically protecting data privacy and trade secrets

Quote documents can contain personal data, floor plans, security information, purchasing conditions, margins, and technical specifics. The Austrian Data Protection Authority clarifies: Whoever decides on the purpose and means of processing remains fundamentally responsible, even if a third-party AI system is used. The GDPR applies in parallel to the AI Act as soon as personal data is processed.

Before use, providers, storage locations, access, retention, deletion, potential use for training, and transfers to third parties must therefore be checked. A free chat account is not a suitable storage location for unfiltered customer files. It is often sufficient to pseudonymize the case: "Project A, Room 3" instead of name and full address; technical dimensions without photos of people; neutral article numbers instead of confidential supplier discounts.

The WKO lists data sovereignty, controlled reuse and auditable deletion as key concerns for privacy-compliant AI in trade and craft businesses. A role-based model also helps: measurements can provide technical data, purchasing current prices, and costing internal markups. Not every person—and not every system—needs access to everything.

AI competence is part of the quality of cost estimates

Since February 2025, Article 4 of the AI Act requires measures to support the AI competencies of people who develop, operate, or deploy AI systems on behalf of a company. Austria’s AI service desk at the RTR stresses that content and depth depend on the system, the risk, prior knowledge and the deployment context. For a craft business, what is needed is not an abstract technical course but concrete practical confidence.

Employees should be able to tell which data may be entered, what hallucinations look like, why source versions matter, and who approves. A short training using the actual quoting process is often more effective than a general presentation. The purpose, approved tools, red lines, sample checks and contact persons should be documented.

AI competence also means seizing opportunities: recurring validation rules, clearer readability and consistent clarifying questions can improve the quality of quotations. It does not mean trusting every suggestion. Being able to object on technical grounds and discard a draft is a sign of competence.

Ten checks to make before sending

Can every quantity be derived from a confirmed measurement or a specified assumption?
Are units, rounding, VAT and calculation methods correct?
Are material prices current and validity periods taken into account?
Do time estimates include setup, transport, assembly, cleaning, and documentation?
Have overheads, subcontracted services, contingencies, and profit been deliberately addressed?
Are the scope of services, options, exclusions, and obligations to cooperate clear?
Has an estimate inadvertently become a firm commitment?
Have technical statements been reviewed by a qualified person?
Have personal data and trade secrets been limited to what is necessary?
Is it documented who commercially approved which version?

A good checklist is short enough for everyday use and strict enough for critical cases. Additional checks can be tied to thresholds: unusually high material proportions, new suppliers, tight deadlines, or lump-sum charges without a full breakdown trigger a second approval.

Which metrics really evaluate a pilot

A four-week pilot should not only measure how quickly a text is produced. More meaningful are the number of missing required fields, corrections before sending, follow-up questions after sending, deviations between pre- and post-calculation, and the time until subject-matter approval. Abandoned drafts are also valuable data: they show where information or rules are missing.

Compare a clearly defined order type in the previous and the AI-assisted process. Do not change pricing logic, software, and responsibilities at the same time. After each week the team discusses two or three cases: Where did the structure help? Where did an error sound convincing? Which input was unnecessarily sensitive? Afterwards the template and checklist are updated and versioned.

Automatically evaluating individual employees based on processing time or number of corrections would be the wrong approach. The goal is a robust process, not covert performance monitoring. Especially experienced professionals find complex errors and sometimes need more time to do so.

The most common false starts

Uploading old quotes unfiltered:This causes names, addresses, prices, and margins to end up with the provider without purpose and contract being clarified. A better option is a sanitized sample set.

Accept the final price without the calculation: A plausible result can conceal multiple incorrect assumptions. Every line item needs its source and formula.

Mistaking a model for master data: AI cannot guarantee an up-to-date supplier list. Prices and availabilities come from authorized systems.

Testing only perfect examples: The pilot must include missing dimensions, conflicting units, and unclear photos. A good system indicates uncertainty.

Leaving responsibility unassigned: If no one is explicitly responsible for technology, cost calculation, and shipping, the neatly worded draft will quickly turn into an unintended approval.

How the process fits into a company AI policy

The proposal process should not be an exception alongside the other rules. Specify in the company AI policy specify which tools are allowed, which data classes exist and when a four-eyes principle applies. The article on "trade secrets in AI tools" helps with building a red list. How results from numbers and tables are checked is shown by the guide Checking AI in Excel.

For each approved use case, a short process sheet is sufficient: purpose, inputs, prohibited data, data source, computation logic, review roles, storage location, deletion deadline, and escalation path. If a new model or a new feature is added, that approval is reviewed again.

Conclusion: Structure quickly, decide slowly enough

AI can noticeably simplify quote calculation in the trades by organizing information, making gaps visible, and clearly phrasing confirmed line items. It becomes a risk when old prices, estimates, or sensitive customer data flow unchecked into an apparently final total price.

The secure Austrian practical path is clear: start with an order type, maintain your own cost data, minimize inputs, label assumptions, make calculation methods transparent, and have every technical and commercial decision approved by a human. That way AI remains strong where it is useful — in structuring and checking — and the company retains cost transparency, customer trust, and responsibility.

Sources and further information

Note: The featured image is an AI-generated illustrative image of a fictional Austrian craft workshop. This post does not replace legal, tax, standards, or business-management advice in individual cases.