Sales & leads
Catering inquiry to proposal in an hour
An event inquiry gets a tailored menu proposal with pricing, staffing and a deposit link while the customer is still comparing options.
Starts when
A catering or private-event inquiry arrives.
What happens
- Read the date, headcount, budget and dietary needs from the inquiry
- Check the calendar and kitchen capacity for that date
- Build a menu and price from your catering packages and current costs
- Draft the proposal with staffing, rentals and deposit terms
In more detail
An automation like this reads a catering or private-event inquiry from email or a web form and pulls out the date, headcount, budget, service style and dietary needs. It checks the event calendar for conflicts and a kitchen capacity rule, then builds a draft proposal from the business's own written packages: menu, per-person price, staffing, rentals, tax and deposit terms. The arithmetic is ordinary code, not the AI. The AI writes the covering email and lists anything it could not match, such as a dietary request the menu does not cover. The events manager reviews every proposal, edits it and sends it, and creates any deposit link. It needs current written packages and prices, one reliable event calendar and clear deposit rules. Businesses already using Tripleseat or Perfect Venue may find those tools cover much of this, and Fenrir Labs could design around them.
You stay in control
The events manager reviews the proposal before it is sent.
Connects to
- Website form
- Email
- Event calendar
- Payments
Questions we would ask you
- Where do inquiries arrive today (web form, email, phone, Instagram, The Knot or similar), and roughly how many per week in peak season?
- Do you use an events system (Tripleseat, Perfect Venue, Caterease) or a shared calendar? Who keeps it current?
- Are your catering packages and prices written down, and when were they last updated?
- How do you decide whether the kitchen can take another event on a given day?
- What deposit and cancellation terms do you use, and how do you collect deposits now?
Discuss this further with the Live Agent
Finance & billing
Supplier invoices to live food costs
Invoices are read line by line, ingredient prices update automatically, and you are told which dishes have slipped below your target margin.
Starts when
A supplier invoice is photographed or emailed.
What happens
- Read every line, including pack sizes and substitutions
- Update the cost of each ingredient
- Recalculate the cost of every recipe that uses it
- Alert you when a dish falls below the margin you set
In more detail
An automation like this reads supplier invoices that arrive as PDFs or photos, turns each line into an item, pack size and price, and updates the cost of that ingredient in the recipe file. Ordinary code then recalculates every recipe that uses the ingredient and flags dishes whose food-cost percentage has moved past the owner's target, with the reason. The chef or owner decides on any menu or price change, and nothing is written to the accounting system without a person. It needs a recipe file with quantities and yields, a food-cost target and invoices that are legible. Reading photographed paper invoices accurately is the hard part. Products such as MarginEdge, xtraCHEF and MarketMan already do this job well, often with direct supplier feeds, so for most restaurants buying one of them is the better first step, and Fenrir Labs could design reporting on top.
You stay in control
The chef or owner decides on any menu or price change.
Connects to
- Email
- Recipe costing
- Accounting
- Point of sale
Questions we would ask you
- How many invoices a week, from how many suppliers, and what share arrive as PDF versus paper?
- Do you already use MarginEdge, xtraCHEF, MarketMan or Restaurant365? If so, what is missing?
- Is there a recipe file with ingredient quantities and yields, and who maintains it?
- What food-cost target do you use, and per dish or per category?
- Who enters invoices into QuickBooks today, and how far behind does it run?
Discuss this further with the Live Agent
Hiring & HR
Staff schedules drafted from the forecast
Next week's schedule is drafted from sales history, reservations, local events and each person's availability, within your labor budget.
Starts when
The weekly scheduling day.
What happens
- Forecast covers for each shift from history, reservations and events
- Work out the staffing each shift needs by role
- Assign shifts using availability, time-off requests and overtime rules
- Show the labor cost against budget before publishing
In more detail
An automation like this drafts next week's staff schedule. It reads sales or cover history from the point-of-sale system, the reservations and private events already booked, and each person's availability and time off from the scheduling tool. Ordinary code forecasts covers per shift, turns them into staffing by role using the manager's own ratios, fills shifts while avoiding overtime, and shows labor cost against the budget. The AI only writes a short explanation of why a shift looks busy. The manager edits and publishes; nothing goes to staff automatically, and decisions about cutting hours stay with the manager. It needs a few months of sales history, written staffing ratios, a labor budget and availability kept in the scheduling tool. Tools such as Homebase, 7shifts and Square already schedule well, so Fenrir Labs could design a forecast that feeds them rather than replace them.
You stay in control
The manager edits and publishes the schedule.
Connects to
- Point of sale
- Reservations
- Scheduling software
Questions we would ask you
- Which POS and scheduling tool do you use, and on which plans?
- How do you build the schedule today: copy last week, a template, or from scratch? How long does it take?
- Do you have written staffing ratios by role and daypart, or is it judgement?
- Where do reservations live, and can you export next week's covers?
- What is your weekly labor budget, and how do you track it now?
Discuss this further with the Live Agent
Customer service
Guest messages answered around the clock
Questions about hours, allergens, parking, private rooms and reservations are answered from your own information on every channel, at any hour.
Starts when
A message arrives by web chat, text, email or social media.
What happens
- Answer from your menu, allergen matrix, policies and calendar
- Offer available reservation times and hold a table
- Take the details for large parties and events
- Hand complaints and anything uncertain to the manager on duty
In more detail
An automation like this answers guest questions about hours, parking, menus, allergens, private rooms and reservations from the business's own approved fact file, across web chat, text, email and social media. The AI classifies the question and phrases the answer, but the facts come from the file, not from the AI. Allergen answers quote the restaurant's allergen matrix exactly and tell the guest to confirm with staff. For bookings it usually sends the reservation link, because reservation systems such as OpenTable and Resy only allow approved partners to book directly. Complaints and anything uncertain go to the manager on duty. A cautious version keeps every reply as a draft that staff send; fully automatic replies raise texting-consent and accuracy questions. It needs a current fact file, an allergen matrix with an owner and a manager watching escalations. Fenrir Labs could design either version.
You stay in control
You approve the source information; allergen answers quote your matrix and advise confirming with staff.
Connects to
- Website chat
- SMS
- Social media
- Reservations
Questions we would ask you
- Which channels bring the most questions: website, Google, Instagram, texts, phone?
- Which ten questions come up most? Can you show a week of real messages?
- Is there an allergen matrix, who updates it, and how often does the menu change?
- Which reservation system do you use, and is a booking link enough?
- Who is the manager on duty, and how should urgent escalations reach them during service?
Discuss this further with the Live Agent
Operations & inventory
Prep lists and orders from pars and reservations
Each afternoon the kitchen gets tomorrow's prep list and a supplier order, calculated from par levels, the reservation count and the events on the calendar.
Starts when
Daily, after the lunch count.
What happens
- Forecast tomorrow's volume from reservations, events and history
- Scale the prep list by station
- Compare stock counts with pars and what the forecast needs
- Draft the order for each supplier before their cut-off time
In more detail
An automation like this prepares tomorrow's prep list and supplier order drafts each afternoon. It reads tomorrow's reservations and private events, recent sales by dish, the latest stock counts and the kitchen's par levels. Ordinary code scales the standard prep list by station, works out what to order from each supplier in case sizes and checks each supplier's cut-off time. There is little AI work here; at most the AI tidies the prep sheet. The chef adjusts the quantities and places every order personally in the supplier's portal or by phone. It needs written pars, regular counts, recipes with yields and a list of suppliers with delivery days. Kitchens willing to count in an inventory app may be better served by a product such as MarketMan, and Fenrir Labs could design a simpler spreadsheet-based version for kitchens that are not.
You stay in control
The chef adjusts and sends each order.
Connects to
- Reservations
- Inventory counts
- Recipe costing
- Supplier ordering
Questions we would ask you
- How do you build the prep list today, and who does it?
- Do you have written pars, and when were they last reviewed?
- How often do you count, which items, and where do counts go?
- Which suppliers do you order from, on which days, and what are their cut-off times?
- Do you order through MOXe, Sysco Shop, another portal, text or phone?
Discuss this further with the Live Agent
Marketing & reviews
Reviews answered and patterns reported
Every review gets a considered reply drafted in your voice, and a monthly summary shows what guests keep praising and what keeps going wrong.
Starts when
A new review is posted.
What happens
- Collect reviews from map listings, reservation sites and delivery platforms
- Draft a reply that refers to the guest's actual visit
- Tag each review by dish, service, wait time and shift
- Report the patterns each month with examples
In more detail
An automation like this drafts a reply to each new review in the owner's voice, referring to what the guest actually wrote, and tags every review by dish, service, wait time and shift. Once a month it produces a short report of recurring praise and complaints, with examples. A manager approves every public reply before it is posted; nothing is posted on its own. Google allows approved businesses to read and reply to reviews through its interface for software, Yelp's paid data service only returns review excerpts and was not found to allow replies, and Tripadvisor has moved to a newer partner platform, so coverage differs by site. The automation never writes, buys or asks for reviews, which federal rules on fake reviews restrict. It needs owner access to the review profiles and a short reply style guide. Fenrir Labs could design a Google-first version.
You stay in control
A manager approves each public reply.
Connects to
- Review sites
- Reservations
- Point of sale
Questions we would ask you
- Where do most reviews appear for you, and roughly how many a month?
- Who replies today, and how quickly?
- Do you have a reply style you like? Can you share replies you are proud of?
- Is your Google Business Profile verified, active for more than 60 days and linked to your website?
- Do you want the monthly report to break down by shift or by server?
Discuss this further with the Live Agent
Hiring & HR
New-hire paperwork and training on a checklist
From the offer onward, forms, certifications, uniform sizes and training shifts are collected and scheduled without a manager chasing each step.
Starts when
A candidate accepts an offer.
What happens
- Send the forms and collect identification and certifications
- Check food-handler and alcohol-service certificates and their expiry dates
- Schedule training shifts with an assigned trainer
- Remind the manager of anything incomplete before the first shift
In more detail
An automation like this runs a checklist for each new hire from the moment they accept an offer. It tracks tax and payroll forms, the handbook acknowledgement, uniform sizes and the certificates the role needs, such as a Texas food handler certificate and a TABC seller-server certificate. It reads uploaded certificates to record expiry dates, proposes training shifts with a trainer for the manager to confirm, and lists anything still missing before the first shift. The manager examines identity and work documents in person for Form I-9, which is due within three business days of the start date; the automation only reminds them. It needs role checklists, an onboarding or payroll tool for the forms, and a secure place for records. Tools such as Homebase already include onboarding, so Fenrir Labs could design a certificate tracker and checklist around them.
You stay in control
The manager verifies original documents in person as the law requires.
Connects to
- E-signature
- Payroll
- Scheduling software
- Email
Questions we would ask you
- How many people do you hire a year, and in which roles?
- Which tool handles W-4, direct deposit and the handbook today?
- Who completes I-9s, and has anything ever been missed?
- Do you use E-Verify?
- How do you track food handler and TABC certificates and their expiry dates now?
Discuss this further with the Live Agent
Documents & compliance
Inspection-ready food safety records
Temperature checks, cleaning tasks and corrective actions are logged from phones and sensors, with gaps flagged the same day rather than at inspection.
Starts when
Scheduled checks through each shift.
What happens
- Prompt the right person for each check at the right time
- Record readings from sensors or a photo of the thermometer
- Flag out-of-range readings and require a corrective action
- Keep a complete, dated record ready to show an inspector
In more detail
An automation like this keeps kitchen food safety records complete. It prompts the person on shift for each scheduled check, such as cooler and hot-holding temperatures, cooling logs, sanitizer strength and closing tasks, and records readings from wireless sensors or typed values. A reading outside the limit, such as a cooler above 41 degrees Fahrenheit, is flagged at once and needs a corrective action written by staff. The kitchen manager reviews and signs off each day's log, and dated records are kept ready for an inspector. There is little AI involved; the value is in reminders, rules and a record that cannot be quietly edited. It needs a written list of checks and limits, calibrated thermometers and staff who use a phone or tablet on the line. Products such as Mitti, formerly SafetyCulture, Jolt, ComplianceMate and Squadle already do this, so Fenrir Labs could design reporting on top of them.
You stay in control
The kitchen manager signs off each day's log.
Connects to
- Mobile phone
- Temperature sensors
- Document storage
Questions we would ask you
- Which checks do you run today, on paper or in an app, and how often are they missed?
- Who inspects you (City of Houston, Harris County, DSHS, another city), and what did the last inspection note?
- Do you use time as a public health control for any food, or any special process such as sous vide or reduced-oxygen packaging?
- Would you install wireless sensors in coolers, and who would maintain them?
- Who is the certified food manager on each shift?
Discuss this further with the Live Agent
