Opportunities › AI customer engagement for professional services › Proposed app
Briefdesk
An AI front desk for small consultancies that answers prospects and logs every reply
- Score
- 33
- confidence 57
- Evidence
- 7
- items on the page
- Rivals
- 2
- named in sources
- First version
- 35 d
- estimate
- Type
- Mobile app
- PWA: installs on iPhone and Android
- Market since
- 2017
- Established market · AI estimate
ScreensConcept
Mobile app: Principals review incoming lead briefs, check replies logged by the AI, and manage front-desk conversations on the go between client meetings. Sample screens on a phone, in the proposed brand. Their numbers are sample data.
Chat preview
- VisitorHomeowner
- QuestionSmall extensions?
- SourceServices, para 2
- Scope checkWithin scope
- HandoffNot needed
About this appAI proposal
Briefdesk replaces the static website of a small architecture, engineering or design practice with a front desk that answers visitors. It handles service questions, repeat FAQs and first qualifying conversations at any hour, so the principal stops retyping the same email replies.
It answers only from material you approve: your services, fees, process and scope limits. When a question falls outside that, it says so and hands the visitor to you with a short summary. It does not guess at code, liability or fee questions.
Every conversation is saved with the source passage behind each answer. You can review, correct and publish transcripts. Tone adapts to who is asking, such as a homeowner or a fellow professional, using settings you control.
- Tagline
- Your front desk, always on, never guessing.
- Voice
- PlainCarefulCalm
- Why this name
- A brief is the document a practice works from, and a desk is where visitors are met. The name says someone attentive answers, from the facts.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Answers only from approved sources
The agent replies from your own services, fee and process notes. If the answer isn't there, it declines and offers a handoff.
No member yet
Audit trail for every reply
Each answer is stored with the passage it came from and a timestamp, so you can show what was said and why.
No member yet
Guardrail topics
Mark subjects such as code compliance, quotes or legal advice as never answered. The agent redirects to you.
No member yet
Qualifying conversations
Asks about project type, location, budget range and timing, then sends you a one-page brief of the lead.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#1F5C4DDrafting Green · primary#F6F1E7Paper · surface#1F2A2EInk · text#B8552FTerracotta · accent#3E6D9CSurvey Blue · data
Typography
Aa
Briefdesk
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
Your front desk, always on, never guessing.
PricingAI proposal
The material gives one price: Wix at $40 a month for a site that cannot answer questions. Solo sits at that level so it can replace the website plan. Higher tiers are a starting point to test.
Solo
$39 a month
- One site and chat agent
- Up to 150 conversations a month
- Audit log and guardrail topics
- Lead briefs by email
Studio
$79 a month
- Up to 600 conversations a month
- Tone profiles per audience
- Review queue and corrections
- Custom domain
- Up to 5 team members
Practice
$149 a month
- Up to 2,000 conversations a month
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$194,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Shared outputs. Visitors see a 'Answered by Briefdesk' line on the chat page, and the lead brief and published transcripts link back to it. Peers and prospects who read them see how the firm works.
- Do they become customers?
- Visitors are mainly the firm's clients, who rarely buy this. Peer professionals who chat with a firm's agent might become customers. Test whether the footer link gets any signups from that group.
- First customers
- Solo and two-to-five person architecture practices in one country, where the principal handles sales and replies to inquiries personally.
- Where to reach them
- Architect and engineer forums: Share a demo agent trained on a sample practice and ask owners to try to trip it up.
- Direct outreach: Email principals whose sites have a contact form only, with a working agent built from their public pages.
- Hacker News and Show HN: Post the guardrail and audit-log design and invite feedback from the builder community.
- First milestone
- Ten small architecture or engineering practices pay for two months, and at least half keep the agent live.
Why people already pay for this
- VerifiedHacker News post received 20 points and 43 comments
- AI inferenceSmall consultancy owner paid $40/month for a website that could not answer customer questions, forcing the principal to waste hours on repetitive FAQs
- Public claimAn AI agent successfully handled complex professional interactions including defending the business model to a licensed architect through sophisticated argumentation
- Public claimPrincipal spent 4 months building an AI agent system despite having not coded in 30 years, indicating strong demand-pull for this capability
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · AI modelHosted LLM API with retrieval over approved notesAnswers must be grounded in the firm's own text, with the source passage stored per reply.
- This app · Embeddings and searchpgvector in Supabase PostgresKeeps approved notes and their passage links in the existing database.
RisksAI inference
- AIThe agent may still give a wrong or overconfident answer. In liability-sensitive fields one bad reply could hurt a firm, so refusal rules and tests need to come before launch.
- CommercialEvidence suggests visitors are mostly one-off inquiries. Owners may see too little value to pay monthly unless the agent clearly saves time or wins leads.
- RegulatoryFirms may be held to what their agent says, and professional bodies may have rules on advice and advertising. Terms and disclaimers need legal review.
- TechnicalTone tuning between homeowners and professional peers reportedly took months for one builder. Defaults that work with little setup may be hard.
- DataSetup needs good source notes from the firm. Thin or outdated notes lead to many handoffs and a weak first impression.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.