Scheduling & dispatch
After-hours call to scheduled job
A voicemail and a photo at 7 pm become a quote, a confirmed appointment and a draft invoice before the office opens.
Starts when
A voicemail or text arrives after closing.
What happens
- Transcribe the message and read the unit's model and serial number from the photo
- Match the caller to their customer record, service plan and past visits
- Check warranty coverage, technician calendars and the parts on each truck
- Draft a quote with two appointment slots and a reply to the customer
In more detail
An automation like this picks up voicemails and texts that arrive after closing. It transcribes the message, reads the model and serial number from any photo of the equipment, finds the caller in the field-service system, and gathers their service plan, past visits, technician availability for tomorrow and parts on the trucks. It then drafts a quote using the business's own price book, two appointment options and a short reply to the customer. The owner reviews and edits everything on their phone, and nothing is booked or sent until they approve. Emergencies such as a gas smell can be routed to an on-call person straight away. It needs a phone system that can export voicemail, a field-service system with an API the business is allowed to use, and registered business texting. Fenrir Labs could design this around the tools a company already runs.
You stay in control
The owner approves or edits the quote from their phone before anything is sent.
Connects to
- Phone system
- CRM
- Calendar
- Accounting
Questions we would ask you
- How many after-hours voicemails and texts arrive in a typical peak-season week, and how many are emergencies?
- Which phone system and which FSM do you use, and on which FSM plan or package?
- Do you already use an answering service or the FSM's AI call answering?
- Who approves quotes now, and from where (phone, laptop)?
- How complete are equipment records (model, serial, install date) in your FSM?
Discuss this further with the Live Agent
Sales & leads
Estimates from customer photos
Customers send photos of the job. They get the right follow-up questions in minutes, and you get a draft estimate built from your own price list.
Starts when
A customer submits photos through your website or by text.
What happens
- Identify the job type and what the photos do and do not show
- Ask the customer only for the missing details, such as access, age or measurements
- Build a line-item estimate from your price list and similar past jobs
- Flag anything that needs a site visit before a firm price
In more detail
An automation like this takes the photos and description a customer sends through the website or by text. A vision model works out what job it is and what the photos show, and a checklist for that trade lists what is still missing, such as unit size, age or access. It drafts a short message asking only for those details, which a person sends. When the answers come back, it builds a line-item estimate using only the company's own price book, shows similar past jobs for comparison, and flags work that needs a site visit before a firm price. An estimator reviews the numbers and sends the estimate. It needs an up-to-date price book, a way for customers to send photos, and registered business texting if questions go by SMS. Fenrir Labs could design the version that fits a company's trade and price book.
You stay in control
An estimator reviews the numbers and sends the estimate.
Connects to
- Website form
- SMS
- Price list
- Estimating software
Questions we would ask you
- Which job types do customers most often send photos for, and which ones do you already quote without a visit?
- Where does your price book live, and when was it last updated?
- How do customers send photos today: text, email, website form, Facebook messages?
- Who builds estimates now, and how long does a typical one take from first contact?
- Would you rather send a range, a firm price, or a "from" price after photos?
Discuss this further with the Live Agent
Scheduling & dispatch
Dispatch that re-plans when a job runs long
When a morning job overruns, the afternoon is reshuffled, the affected customers get a new window and the technician's route updates.
Starts when
A technician marks a job as running over, or GPS shows they are still on site.
What happens
- Work out which later appointments are now at risk
- Find the nearest technician with the right skills and parts
- Propose a revised schedule that respects promised time windows
- Draft a text to each affected customer with their new arrival time
In more detail
An automation like this watches the day's schedule. When a technician marks a job as running over, or the truck has been at the address much longer than planned, it works out which later appointments will now miss their promised window. A routing optimiser looks for another technician with the right skills and parts nearby and proposes a revised afternoon. A language model writes a short explanation for the dispatcher and a draft text for each affected customer with the new arrival time. The dispatcher accepts with one click or changes it, and only then are appointments moved and texts sent. It needs a field-service system with live events and an API the business can use, skills and durations recorded, and ideally GPS. ServiceTitan users should look at its own dispatch add-on first. Fenrir Labs could design a version for smaller shops on other systems.
You stay in control
The dispatcher accepts the new plan with one click or adjusts it.
Connects to
- Field service software
- Calendar
- GPS
- SMS
Questions we would ask you
- How often does a morning overrun push afternoon jobs, in a typical peak week?
- Who dispatches, and how do they find out a job is running long today?
- Which FSM and which plan or package are you on, and do you use its dispatch add-ons?
- Do your trucks have GPS, and which provider?
- Are skills and truck inventory recorded in the FSM and kept current?
Discuss this further with the Live Agent
Finance & billing
Voice notes to finished invoice
The technician talks through what they did on the drive back. The invoice, the job record and the customer summary are written for them.
Starts when
A technician records a voice note when the job is done.
What happens
- Turn the note into labor, parts and materials line items
- Price each line from your rate sheet and the customer's plan discounts
- Write a plain-language summary of the work for the customer
- Update the equipment history on the customer record
In more detail
An automation like this lets a technician talk through the job on a voice note after parking. The note is transcribed, and a language model pulls out the tasks done, parts used, time on site and anything the technician recommended. Plain code then prices each line from the company's own price book and applies the customer's plan discount, so the model never makes up a price. It also writes a short summary of the work for the customer and proposes updates to the equipment history. The office checks the draft invoice, corrects anything wrong and sends it from the field-service system. It needs a current price book, customer plan records and technicians willing to record a note per job. Fenrir Labs could design this to run on the company's own computer and fit its field-service system.
You stay in control
The office checks the draft invoice before it goes out.
Connects to
- Mobile phone
- Field service software
- Accounting
Questions we would ask you
- How long after a job is the invoice sent today, and who writes it?
- Do technicians already write notes in the FSM app, and how complete are they?
- Is your pricing flat-rate from a price book, time and materials, or a mix?
- How many price-book items do you have, and how often do technicians use items not in it?
- Which FSM and plan are you on, and does it sync to QuickBooks?
Discuss this further with the Live Agent
Sales & leads
Maintenance-plan renewals that run themselves
Every plan coming up for renewal gets a personal message that mentions the customer's own equipment and last visit, not a generic reminder.
Starts when
A service plan reaches 45 days before expiry.
What happens
- Pull the customer's equipment, visit history and plan usage
- Write a renewal message that refers to what was actually done this year
- Send by the customer's preferred channel and watch for a reply
- Follow up twice, then hand unanswered renewals to the office with a call list
In more detail
An automation like this looks at maintenance plans coming up for renewal, about 45 days out. For each customer it reads the equipment, this year's visits and what was found, and drafts a renewal message that mentions those specifics, with price and terms filled in from the plan record rather than written by the model. It checks consent and opt-outs before choosing text or email, and the office releases each batch. Replies are sorted into renewals, questions and stop requests, and stop requests are honoured straight away. Unanswered renewals go to the office as a call list after two follow-ups. It needs agreements recorded in the field-service system, consent records for texting, and clear renewal terms. Texting rules such as the TCPA and Texas law apply, so a lawyer should review the setup. Fenrir Labs could design this around the systems a company already uses.
You stay in control
You approve the message template and any price change once per season.
Connects to
Questions we would ask you
- How many active maintenance agreements do you have, and what share renew today?
- Where are agreements tracked, and are renewal dates and terms in the FSM?
- Do your plans auto-renew with a card on file, or does the customer re-buy?
- What consent do you hold for texts and emails, and where is it recorded?
- Who follows up on renewals now, and how?
Discuss this further with the Live Agent
Operations & inventory
Truck stock that reorders itself
Parts used on jobs come off each truck's inventory automatically, and a purchase order is ready when stock runs low.
Starts when
A job is closed with parts recorded.
What happens
- Deduct the parts used from that truck's stock
- Compare stock against the minimum you set for each part
- Group what is needed into one order per supplier at your agreed prices
- Note which upcoming jobs depend on a part that is short
In more detail
An automation like this keeps a running count of parts on each truck. When a job is closed, the parts recorded on it come off that truck's stock. Each night it compares stock with the minimum levels the owner set, groups what is short into one draft order per supplier at the company's agreed prices, and lists upcoming jobs that need a part the truck does not have. The purchasing lead reviews and approves each order and places it through the supplier's website or by email; nothing is ordered automatically. Most of this is ordinary software rather than AI. It only works if technicians record the parts they use, trucks are counted from time to time, and the company has supplier price lists. ServiceTitan users already have replenishment built in. Fenrir Labs could design a lighter version for shops on other systems.
You stay in control
The purchasing lead approves each order before it is placed.
Connects to
- Field service software
- Inventory
- Supplier portal
- Accounting
Questions we would ask you
- Do technicians record parts on every job today, and how accurately?
- Do you track stock per truck now, in the FSM, a spreadsheet or not at all?
- Which suppliers do you buy from, and how do you order (portal, phone, counter)?
- Do you have agreed price lists from suppliers, and in what format?
- Who approves purchases, and is there a spending limit?
Discuss this further with the Live Agent
Documents & compliance
Permit and inspection tracker
Every job that needs a permit has one applied for, tracked and closed out, with inspection dates on the calendar and the customer kept informed.
Starts when
A job is sold that falls under a permit rule you define.
What happens
- Fill the application from the job record and equipment specifications
- Track the status and read emails from the permit office
- Put inspections on the calendar and make sure a technician is assigned
- File the signed-off permit with the job and tell the customer
In more detail
An automation like this keeps permits from slipping. When a job is sold that matches a permit rule the owner sets for each city, it opens a permit record and drafts the application from the job and equipment details. A licensed person checks, signs and files it on the city's portal; the automation does not file anything itself. Afterwards it reads emails from the permit office, keeps the status on the job up to date, makes sure each required inspection has a date and a technician, and lists jobs still missing a final inspection. When the permit is closed it files the paperwork with the job and drafts a note to the customer. It needs a list of permit rules per jurisdiction, permit emails going to a shared inbox, and equipment details on each job. Fenrir Labs could design a tracker for a company's own cities.
You stay in control
A licensed person signs each application before it is filed.
Connects to
- Job records
- Email
- Calendar
- Document storage
Questions we would ask you
- Which cities and counties do you pull permits in, and roughly how many per month?
- Who pulls permits now, and how do they know a job needs one?
- Where do permit emails go, and who reads them?
- How do you schedule inspections in each jurisdiction (phone, portal, email)?
- How often does a job close without a final inspection, and how do you find out?
Discuss this further with the Live Agent
Marketing & reviews
Seasonal tune-up campaign by equipment age
Before the busy season, each customer whose system is due gets an offer that fits their equipment, sent in waves that match your crew's capacity.
Starts when
The date you set for the spring or fall campaign.
What happens
- Select customers by equipment age, last service date and location
- Write an offer specific to each customer's system
- Release messages in batches sized to the open slots on the calendar
- Let customers pick a slot, then add the visit and confirm it
In more detail
An automation like this runs the spring or fall tune-up campaign from the business's own customer records. On the date the owner chooses, it selects customers whose equipment is due by age, last service date and area, removes anyone without consent or who has opted out, and drafts an offer that fits each customer's system, with prices filled in from the owner's settings rather than written by the model. Messages go out in batches sized to the open slots on the calendar, so demand matches crew capacity. The owner approves the offer and the audience before the first batch, and the office releases later batches and confirms bookings. It needs equipment ages in the field-service system, consent records for texting, and online booking or an office ready to confirm. Texting rules such as the TCPA and Texas law apply. Fenrir Labs could design this around a company's systems.
You stay in control
You approve the offer and the audience before the first batch goes out.
Connects to
Questions we would ask you
- How many customers have equipment with a known install year in your FSM?
- What consent do you hold for marketing texts and emails, and where is it recorded?
- How many tune-up slots can your crew handle per week in spring and fall?
- Which FSM and package are you on, and do you use its marketing add-on?
- Do you offer online booking today?
Discuss this further with the Live Agent
