
Superintelligence for Drain Cleaning Companies: Triage, Quotes, On-Call
How to triage urgent drain clearing calls, send a standard quote range and dispatch the on-call technician without promising a fixed price before the visit.
Want to apply this in your company?
Book a call with Juan →In a drain cleaning company, useful AI does not set prices or promise deadlines. It organizes the call: it understands the problem, separates a real emergency from a routine job, presents a standard quote range with written conditions, and dispatches the right technician from the on-call schedule. The term "superintelligence" works here as a search term. What you actually operate is an agent with clear rules, defined limits and a record of every decision.
Key takeaways
- Triage comes before any promise: the agent collects symptoms, location, property type and access, then classifies urgency using rules the company has written.
- A standard quote is a range with conditions, not a fixed price: the technical visit confirms the scope before the final price.
- On-call dispatch needs a defined schedule, clear escalation criteria and a human handoff with a complete case summary.
- Every call leaves a record: what was asked, what was told, which range was presented and who made the decision.
- Start with a small flow, measure the time to first human contact, and expand only if the results justify it.
Why a drain call needs triage
A customer writes "sink is blocked," but the real problem could be grease buildup in the kitchen, a clogged utility sink, a toilet at risk of overflowing, or sewage backing up at more than one point in the home. Each case has different urgency, complexity and cost. Without triage, the crew drives to the site with too little information, or someone gives an estimate over the phone based on guesswork.
The problem repeats on the sales side. When a representative improvises a number, the customer treats it as a promise. When nobody answers, the customer calls a competitor. A well-designed triage solves both extremes: it asks the right questions, shares what can responsibly be shared, and passes the rest to someone who can decide.
How to build the triage in steps
The sequence below is a starting point. Your technical team should validate each step before it goes live.
- Identify the type of service and confirm the address only to the level needed for the visit.
- Ask about the main symptom: slow drain, toilet that won't flush, water backing up, odor, flooding.
- Record how many fixtures are affected and whether the problem started suddenly or gradually.
- Check site access: inspection box, backyard, upper floor, and when someone will be present.
- Classify urgency using the company's table: routine, priority or emergency.
- Present the approved standard quote range, along with the conditions that could change it.
- Confirm the visit window, or dispatch on-call when the classification requires it.
- Save a summary of the call for the technician so the customer does not have to repeat the questions.
The list shows the order, not the tone. The agent should be direct, make it clear when it is collecting information, and never pretend it has already diagnosed the problem.
How to share a standard quote without locking in a price
The standard quote is a service table approved by the company itself. The AI can present that table, explain what is included, and list what could change the amount: a larger scope than described, extra materials, difficult access, after-hours dispatch, or the need for specialized equipment. The final price exists only after the visit or after human confirmation.
This protects both sides. The customer understands the general range and the conditions behind it. The company avoids committing to a price that depends on something only the technician can see on site. The owner or the commercial lead defines the values. The agent only presents what has been approved and does not negotiate discounts or create exceptions.
Illustrative scenario: a customer describes a toilet that drains slowly and asks what it costs. The agent confirms the type of service, explains that standard clearing has a defined range in the table, notes that the amount may change if a fixture must be removed or heavier equipment is needed, and schedules the visit. It does not state a fixed final price and does not promise the problem will be solved in a single trip.
When to dispatch the on-call technician
Not every urgent call needs on-call dispatch. The decision should follow written criteria, so the agent does not dispatch on a hunch and the team does not get off-hours calls that were unnecessary. A simple decision guide usually includes these criteria:
- Sewage coming back into the property or rising through a bathroom drain.
- Toilet water at a high level with overflow risk, especially when people are home.
- Flooding or seepage into a lower floor with damage already in progress.
- Gas smell or suspected leak: the agent directs the customer to the utility's or the building's safety procedure and does not dispatch drain clearing.
- Off-hours call with no sign of risk: log it and schedule for the next shift.
- Customer asks to speak with a person: transfer immediately, with the summary.
The handoff must be complete. The technician or representative should receive the address, the symptom, the classification, the quote range already presented and what the customer has confirmed. If the customer has to repeat everything, the process has failed.
Limits the agent must not cross
Some limits should be written down before any configuration. The agent does not promise an arrival time that depends on route and availability, does not guarantee that clearing will fix the cause, does not make a definitive technical diagnosis from a description, does not close a job without confirmation, and does not change prices. It also should not accept instructions a customer types to change its behavior, such as a request to ignore the price table. OWASP documents this type of risk in applications built on language models, and the practical rule is to treat user messages as data, not as commands.
On personal data, the principle is to collect only what the visit requires. Address, phone number and a description of the problem are usually enough. If customers send photos, the company should define how long it keeps them and who can access them. To understand the obligations around collection, purpose and data subject rights, consult the National Data Protection Authority (ANPD) and the text of the LGPD.
Decision records and follow-up
A triage agent without records cannot be audited. Every call should store the time of entry, the assigned classification, the quote range presented, the reason for any on-call dispatch and the outcome. That history makes it possible to review cases where a customer complained, adjust rules that classify incorrectly, and find questions the agent consistently leaves unanswered.
Follow-up should be simple and regular. Useful indicators include the time to first human contact, the share of calls sent to on-call without need, how often the quote range had to be corrected after the visit, and the reasons for cancellation. The goal is not to measure the team but to learn whether the rules are working. If you want to see how this kind of agent is kept running over time, the work on AI agents for electronics repair shops compares similar intake flows, and the case of call, visit and warranty handling for air conditioning shows the same questions in another service business.
The risk of treating the whole thing as a black box appears when nobody knows why the agent said what it said. For that reason, the NIST AI Risk Management Framework recommends managing AI risk with measurement, defined responsibilities and continuous review, and that logic fits a service intake agent well.
How to start with a pilot
The approach we apply in projects like this has four steps: map the current process and its baseline, limit data and actions, pilot a real flow with records and a success metric agreed in advance, then hand over the operation and decide whether to expand, adjust or stop. In a drain cleaning business, mapping often shows that most of the time is lost to repeated questions and callbacks, not to the visit itself. Starting with one type of call, with an approved price table and a defined on-call rule, is usually more useful than trying to automate the whole company at once.
If the model is managed by Be Human (be-human.ai), follow-up applies the same principle: what the agent may do, what it is forbidden to do, who receives the handoff and where the record is kept. To see the difference between what the word "superintelligence" promises and what is actually running today, the article on ChatGPT and superintelligence explains the term plainly.
Frequently asked questions
Can the agent give the final price for a drain clearing job over WhatsApp? We do not recommend it. The agent can present the range from the standard table and the conditions that change the amount, but the final price depends on a visit or human confirmation.
When should a call go straight to the on-call technician? When there is a sign of risk, such as sewage backing up, flooding in progress, or a toilet about to overflow with people in the home. The criteria should be written down and validated by the technical team, not decided by the agent on a guess.
Do we need to keep the customer conversation? Yes, but only what the visit and process review require. The company should set the retention period, define who has access and respond to access or correction requests in line with the LGPD.
Should we start with every service the drain company offers? Usually not. Starting with one type of call that has a clear price table and a well-defined on-call rule lets you measure results and adjust the rules before widening the scope.