AI agents for architects that answer the enquiry while the team is on site or at the drawing board, ask the questions you wrote about the project, the site, the budget band and the timeline, answer what a prospective client asks about your work and your process from your own pages, book the initial consultation into your calendar, and keep following up by email for as long as the decision takes. Every enquiry kept, with its source.
7-day free trial · Cancel anytime
Somebody who has spent an evening looking at your work lands on the contact page and opens the chat widget. The agent answers what they ask about the kind of projects you take, how you work and what a first meeting involves, from your own pages and Q&A pairs, and then asks the questions you wrote: what the project is, where the site is, the budget band, when they hope to start, whether they own the site yet. The answers are recorded against that person, and the agent offers a real open slot for an initial consultation if the enquiry fits the criteria you set.
The widget is one line of script, in your colour, with your image and a greeting you wrote. It asks for an email address when the conversation is worth keeping, and that address ties the visitor to the email or call they send later, so nobody is asked the same questions twice.
A small practice cannot answer the phone during a site visit, a client presentation or a planning meeting, and that is when enquiries call. The agent answers on your own Twilio number, in real time, from the script you saved. It tells the caller what the practice does and what a first meeting involves, asks your qualifying questions, and books the consultation from real open slots. Every call is recorded and transcribed and lands in the inbox against the caller's record, tagged as a call, next to the chats and emails.
A caller who is a current client, a contractor about a live project, or a planning officer is one you tell the agent to hand over. The thread comes to your inbox with the agent standing down, and the recording is there to listen to before you ring back. Answering and dialling are separate switches.
A house or a commercial fit-out is decided over months, not days. The prospect who had a good first conversation in March and went quiet is followed up by a sequence you wrote, from your own mailbox: a note a week later, another in a month, a project you have just completed after that. Each step has its own delay and send condition, the whole sequence runs inside the hours you set, and it stops the moment the person replies, books or unsubscribes.
Connect Gmail or Outlook with an app password, or any other mailbox as SMTP and IMAP. Mail goes through your own server, so it arrives from the practice, and each mailbox carries a health score so a follow-up sent in month four still lands in the inbox and not in spam.
Every chat, call transcript, email and DM is stored against the person it belongs to, in order, with timestamps, and every reply the agent wrote is stored alongside the material it was generated from. When a director wants to know what a prospect was told about fees or timing before the first meeting, the thread is there with the page or the Q&A pair it came from.
The dashboard counts people, not messages: how many enquiries the agent reached, how many replied, how many visited your site and which pages they read, and how many booked a consultation, by channel and by agent. Set a goal with a dollar value on a signed appointment and the funnel reports which channel, and which campaign, produced the projects the practice actually took on.
An architecture practice lives on a small number of enquiries, each of them worth a great deal and each of them decided slowly. The problem is not that enquiries are rare. It is that the people who could answer them properly are on site, in a client presentation or at the drawing board, and the enquiry that arrives at 4pm on a Thursday is answered on Monday, briefly, by whoever is at their desk. The prospective client who spent an evening on your portfolio and wrote a careful message gets a two-line reply three days later, and in the meantime has had a long phone call with another practice. AI agents for architects exist to give every enquiry the first conversation it deserves, at the moment it arrives.
Bold Agent Kit gives you one agent that answers every channel a prospective client uses: the chat widget on your website, the phone through your own Twilio number in real time, the studio mailbox, text messages on your own number, and the DMs on Instagram, Messenger, WhatsApp and eight other platforms. You give it what it needs once: your project pages, your process page, your fee-structure and lead-time data points, your Q&A pairs for the questions every new client asks, and the qualifying questions you want asked, in the order you want them. Project type, site, budget band, timeline, whether they own the site, whether they have planning permission. Then it runs the same conversation on every enquiry, in the studio's voice, and books the initial consultation into your calendar when the enquiry fits.
It answers only from what you gave it. What kind of projects the practice takes, how the process runs from feasibility to completion, what a first meeting involves and what it costs if you charge for one, and which of your projects are most like the one being described, all from your own pages and documents, with links only to pages of your site you allowed. It does not quote a fee you did not give it as a data point, and it does not promise a timeline. A question it was not given the answer to, a current client, a contractor about a live job, a planning officer, goes to your inbox with the whole conversation attached and the agent standing down. The practice decides what may be said.
And then it does the thing a small practice never has time for: it keeps following up. A prospect who is deciding over months is written to by email on the schedule you wrote, a note a week later, a completed project a month after that, until they book, reply or unsubscribe. Past clients are written to when you want a referral, and the builders, agents and consultants who send you work are written to on a schedule of their own. Every message goes out from your own mailbox, inside your hours, with a one-click unsubscribe, and every reply is worked by the agent and kept against the person. All of it is $25 a month for the practice, with unlimited agents.
A practice with excellent work still loses projects at the front door, because the front door is watched by people whose real job is somewhere else. Four things get in the way.
The phone rings during a site visit, a client presentation or a planning meeting, and it goes to voicemail. The person on the other end has a plot, a budget and a decision to make about which practice to talk to, and most of them do not leave a message. They ring the next practice whose work they liked, and that practice picks up.
A prospective client writes three paragraphs about the site, the brief and what they have seen in your portfolio. It sits in the studio inbox until somebody is free, and gets two lines back on Monday asking them to call. The questions that would have told you whether it was a real project, and told them you had read the message, are never asked.
A house is decided over a year. The prospect who had a good first meeting in March, and then went quiet while they bought the site, is not written to in April, May or June, because following up is nobody's job in a practice of six. When they are ready in September they go to whichever practice was last in touch.
The Instagram DM about a house in your feed, the chat on the website, the studio mailbox, the director's mobile and the form on a directory site are each watched by a different person, or by nobody. The same person appears in two of them as two strangers, and nobody can say which channel produced last year's projects.
You build the agent once, with your questions, your pages and your hand-over rules. From then on it takes every enquiry on the channels you connect, qualifies it, books the meeting and keeps in touch, and only fetches you for the conversations that need an architect.
It answers the chat widget, the call on your own Twilio number, the studio mailbox and the Instagram DM from your project pages, your process page, your Q&A pairs and your fee and lead-time data points. What kind of work you take, how the stages run, what a first meeting involves: answered. A fee you did not give it, or a timeline: handed to a person.
Project type, site location, budget band, timeline, whether they own the site, whether they have planning. The agent asks, records the answers in the thread against that person and, with HubSpot or GoHighLevel connected, writes them to the record. The enquiry that meets your criteria is offered a consultation; the one that does not gets the reply you wrote for it.
It is given your real open slots for the week ahead, offers a few, and writes the one the client picks into your calendar. The enquiry that is not ready is followed up by email on a schedule you wrote, for as long as you set, and the sequence stops the moment they book, reply or unsubscribe.
A practice's Instagram is often where a prospective client first sees the work, and the DM that follows is an enquiry that nobody in the studio treats as one. Connect Instagram, Messenger, WhatsApp and the other eight platforms with your own app credentials and the same agent answers there, inside each platform's reply window, from the same pages and with the same qualifying questions it asks on the website. A DM asking "do you do extensions like this one?" is answered from your project page and asked where the site is.
Every handle a person uses is attached to one record, so the prospect who sent a DM in January and emailed the studio in March is one person with one history, and the agent reads that history before it replies. Every DM lands in the same inbox as the calls and emails, tagged with the platform, and a person can take the thread over at any point.
Each one answers a problem above, and each one runs without anybody in the studio remembering to do it.
The chat, the call, the email and the DM are answered the moment they arrive, from your pages, with your questions. Not on Monday.
Every enquiry is asked the same questions in the same order, so the studio knows the site, the budget band and the timeline before anyone spends an hour on it.
A note a week later, a completed project a month after that, and on for as long as you set. It stops the moment they reply, book or unsubscribe, and it never forgets one because the studio is busy.
One row per person, every message and call in order, with the source of every reply and the channel they first arrived on, so the practice knows where its projects come from.
Every chat, call, email, DM, follow-up and booking is recorded against a person and reported as a funnel, per agent and across all agents. This is the part that tells the practice whether it is the portfolio site, Instagram or the referral email that produces the projects it takes on.
The funnel counts people rather than messages: how many enquiries the agent reached, how many replied, how many went on to visit your site, and how many booked a consultation, with the total value won and the median time from first contact to booking. Set a goal with a dollar value on a booked consultation, or mark an appointment as won by hand with the fee attached, and the dashboard reports what each channel and each follow-up campaign brought in over a year.
Below the funnel, a channel table shows where each person first made contact and which channel they were on when they booked. A table of pages shows which projects in your portfolio known prospects read before they booked, so you learn which work brings in the enquiries you want. One row per person shows the whole history in order, across months, and, with the one-line snippet on your site, visit triggers close the loop: when a prospect from March comes back to your process page in August, the agent tells you, or follows up after a delay on the channel they last used.
Built from the cookie, email address and phone number attached to that contact, so activity months apart on different channels appears on one row.
Page goals and visit triggers require a one-line snippet on your website. Conversations, bookings and manually marking a deal won work without it.
Each of these is usually a person whose real job is somewhere else, a separate service or a separate tool, and most of the tools are priced per call, per conversation, per contact or per seat. Here they are one plan.
| The job | What it normally takes | Bold Agent Kit |
|---|---|---|
| Answering the phone while the team is on site | An answering service priced per call, or voicemail | Included, on your own number |
| Asking the qualifying questions on every enquiry | Whoever was at their desk, if they thought of it | Included, your questions in your order |
| Answering "do you do projects like this?" and "how do you work?" | An architect, interrupted, from memory | Included, from your pages and Q&A pairs |
| Live chat on the website | A chat vendor, priced per conversation | Included, from your own material |
| Booking the initial consultation | A booking link, and hoping they use it | Included, the agent books it |
| Following up an enquiry for six months | Nobody, in a practice of six | Included, by email on your schedule |
| Instagram and Messenger DMs treated as enquiries | Checked by hand, or not | Included, in the same inbox |
| The AI itself | Metered per minute, per conversation or per credit | Your own key, billed by your provider at cost |
Bold Agent Kit is $25 a month, or $200 a year, for the whole practice: every agent, every channel, every enquiry, no per-call or per-seat charge. The only other costs are your carrier's, for calls and texts, and your AI provider's, typically a few dollars a month.
The same agent covers all of these. Many practices run one for new enquiries and another for past clients and referral partners, or one per studio.
A practice that wants every enquiry answered and qualified the moment it arrives. Connect the agent to the chat widget, the studio line, the mailbox and Instagram, load the portfolio and process pages, and never create a campaign. It answers what comes in, asks your questions, books the consultation for the enquiries that fit, and hands the rest to a person.
A practice with a list of people who enquired, had a first conversation and are still deciding. Add them to a list and run a follow-up sequence you wrote: a note a week later, a completed project a month after that, a check-in the month after. Inside your hours, from your own mailbox, until they book, reply or unsubscribe. The agent answers the replies and books the ones who are ready.
A practice whose best enquiries come from people it has already built for, and from the builders, estate agents and engineers who send work its way. Pull the segment from HubSpot or GoHighLevel, or upload a CSV, and run a campaign at it: a completed project to past clients, a note to referral partners about the kind of work you are taking on. The agent answers the replies.
Connect the tools you already run and the agent works them like an assistant would: it takes appointments straight into your calendar, writes notes and outcomes back to the CRM, and reads your store to see what somebody has bought before it answers them.
Bookings are real, not guesses. It is only ever given slots that are genuinely free, and it can only take one it actually offered — so it cannot invent a time or double-book you. The same connection pulls a segment of past customers into a list when you want to run a campaign at them.
HubSpot
GoHighLevel
Housecall Pro
ServiceTitan
Jobber
Google Calendar
Microsoft 365
Calendly
HubSpot Meetings
Zoho Bookings
Acuity
Cal.com
Square
Stripe
Shopify
WooCommerce
Wix
SquarespaceEach agent uses a model on your own API key, in the same way it sends from your own mailbox and your own phone number. Sixteen providers are supported: OpenAI, Anthropic, Gemini, xAI and Mistral; DeepSeek, GLM, Kimi, Qwen and MiniMax; Groq, Together, Fireworks and OpenRouter; and Ollama or any other OpenAI-compatible endpoint you host yourself.
Bring your own credentials and you own the whole stack: your model provider, your mailboxes, your carrier, your CRM. We charge for the software and nothing else — no token markup, no metering, no credit packs — so your provider bills you direct at their rates, usually a few dollars a month per agent.
OpenAI
Anthropic
Gemini
xAI
Mistral
DeepSeek
GLM
Kimi
Qwen
MiniMax
Groq
Together
Fireworks
OpenRouter
Ollama
Any OpenAI APIA new address sending forty cold emails a day gets filtered, and takes the domain with it. Mailboxes that opt into warm-up send each other ordinary mail on a ramp, starting at five a day. The receiving mailbox finds each message, moves it out of spam if it landed there, marks it read and replies to a share of them.
Each mailbox carries a health score out of 100 — SPF, DKIM and DMARC on the sending domain, seven days of real sends and bounces, and where its own warm-up mail is landing — with the reasons listed. A mailbox whose bounce rate or spam placement passes a threshold is pulled out of campaign rotation automatically, before it damages the domain.
Three steps. Inbound answering is a switch per channel, and a follow-up campaign is created as a draft that shows you the whole plan, every person and every step, before it sends anything.
Name it, choose a model and paste your API key. Then give it what it needs to answer a prospective client: your project pages and your process page, read and stored so it can link to them verbatim; your fee structure and typical lead times as data points, if you want them quoted; your practice profile as a document; and your Q&A pairs for the questions every enquiry asks. "Do you take projects of this size?", "how does the process work?", "what happens at a first meeting?".
Write the instructions last: the qualifying questions in the order you want them asked, the criteria for offering a consultation and the reply for an enquiry that does not meet them, what it may say about fees and timing, and the exact situations in which it must stop and fetch a person. A current client, a contractor about a live project, a planning officer and a fee question you did not give it are the usual four.
Add the chat snippet to your website, the studio line on Twilio, the studio mailbox, and the Instagram and other social accounts where enquiries actually arrive. Each one runs on your own account, so the calls come from your number, the emails from your domain, and the reputation stays yours.
Then add the calendar the consultations go into, Google Calendar, Microsoft 365, Calendly or Cal.com, and the CRM if you run one, so the agent writes the qualifying answers to the record and knows a past client when one writes. For the past-client and referral campaigns, pull the segment from the connected CRM or upload a CSV.
Switch auto-reply on for the channels you want the agent to answer. Answering calls is its own switch, and dialling is another, so the agent can pick up the studio line without ever placing a call. From that moment it takes every enquiry that arrives, on every channel you connected.
For the enquiries still deciding, and for past clients and referral partners, build a campaign, set the days, hours and time zone it may send, and read the whole plan on the review screen: every person, every step, every date over the months ahead. Nothing leaves until you press Start. Then watch the inbox: replies are worked by the agent, anything you named comes to you, and a booking ends the sequence.
One account runs unlimited agents, one per client, and nothing is shared between them. A knowledge base belongs to a single agent, and content from one agent is never used to generate replies for another.
Invite a client by email address and they get their own login, which opens onto the one agent you gave them: its conversations, its numbers, its analytics and a reply box. They cannot see your other clients, your API keys, your settings or your billing. Access is deny-by-default.
No seat tiers, no credit packs, no per-agent or per-channel pricing, and nothing held back behind an enterprise call.
| Agents | |
|---|---|
| How many | Unlimited — one per business, brand or job |
| Shared brain | The same identity, instructions and knowledge on every channel |
| Knowledge base | Documents, Q&A pairs, data points and pages of your own site |
| Isolation | One agent's content is never used to reply for another |
| Client seats | Invite a client into one agent, with their own login |
| Channels | |
| As many mailboxes per agent as you connect, rotated by campaigns | |
| Text messages | Nine providers — Twilio, Telnyx, SimpleTexting, Plivo, Vonage, Sinch, Bandwidth, MessageBird, ClickSend |
| Phone calls | Answering and dialling, with recordings and transcripts |
| Transcription | Google, or Deepgram nova-3 on your own key |
| Voices | Google, Amazon Polly and ElevenLabs, or your own ElevenLabs / Fish Audio account |
| Voice cloning | Record a voice in the browser and clone it on your ElevenLabs or Fish Audio key |
| Website chat | A widget on your site with your colour, image and greeting |
| Social DMs | Messenger, Instagram, WhatsApp, Telegram, LINE, Viber, Slack, Discord, Bluesky, Reddit, X |
| Inbox | |
| One list | Every channel's conversations together, newest first, tagged by channel |
| Filters | By channel, by mailbox or number, or by agent |
| Takeover | Reply yourself and the agent stands down on that thread |
| Hand back | Return the thread to the agent when you are done |
| Campaigns | |
| Multichannel sequences | Email, text and call steps in one campaign, each with its own timing |
| Send conditions | Send a step only if they opened, did not open, clicked or did not click |
| Schedule | The days, hours and time zone anything may be sent |
| Stop rules | A reply, a booking or an opt-out on any channel ends the sequence |
| Pacing | Per-mailbox and per-number hourly limits, pausable and resumable |
| Review before start | The full plan is shown; a campaign sends nothing until you start it |
| Templates | Email, text and call scripts, saved once and reused |
| Analytics | |
| Funnel | Reached, replied, visited, converted, plus value won and time to convert |
| By channel | Where each person started, and the channel they converted on |
| Pages | Which pages known people read, and how many of them converted |
| Journeys | One row per person: every message, visit, call and booking in order |
| Goals | A page rule or a manual "won", each with a dollar value |
| Triggers | A known person returns to a page — notify you, or send a follow-up |
| Open and click tracking | Recorded per campaign and added to that person's history |
| Integrations | |
| Calendars | Google, Microsoft 365, Calendly, Cal.com, Acuity, Zoho, Square, HubSpot Meetings |
| CRMs and commerce | HubSpot, GoHighLevel, Housecall Pro, ServiceTitan, Jobber, Stripe, Shopify, WooCommerce, Wix, Squarespace |
| Booking | Real open slots offered, and only an offered slot can be booked |
| CRM import | Import a segment of customers into a contact list |
| Deliverability and compliance | |
| Mailbox warm-up | Ramped sending across the network, with spam rescue and replies |
| Health score | SPF, DKIM, DMARC, bounce rate and inbox placement, with reasons |
| Auto-rest | A failing address is removed from rotation automatically |
| Unsubscribes | One click, honoured automatically, including on queued mail |
| Reply limits | One reply per thread, per-sender daily and per-channel hourly caps |
| Filtering | Bounces, out-of-office and no-reply addresses ignored |
| List building | |
| Contact extraction | Emails, phone numbers and social profiles, on any site. Unlimited. |
| Deep scan | Follows the pages of a domain most likely to carry contact details |
| Domain queue | Load a CSV of sites and work through them one at a time |
| Contacts and lists | Unlimited, by email or phone, with CSV import and export |
| AI | |
| Your own key | Sixteen providers, including any OpenAI-compatible endpoint you host |
| What we charge for AI | Nothing — no markup, no metering, no credit packs |
| On calls | Uses your provider's fastest model automatically |
| Account | |
| Platforms | Chrome side panel and web dashboard on one login |
| Members' forum | Six boards, visible only once you are signed in |
| Referral program | $50 a year per referral ($5 a month), recurring for the life of the account |
| Support | support@boldagents.co |
Billed monthly. No add-ons, no seat tiers. Cancel anytime.
$16.67 a month, billed once a year. No add-ons, no seat tiers. Cancel anytime.
They have the first conversation the studio has no time for. When an enquiry arrives by chat, phone, email or Instagram DM, the agent answers what the person asks about the practice from your own pages, then asks the qualifying questions you wrote: project type, site, budget band, timeline, ownership, planning status. The answers are recorded against that person and, with a CRM connected, written to the record. If the enquiry meets the criteria you set, the agent offers real open slots for a consultation and books the one they pick; if not, it sends the reply you wrote for that case. Anything you told it to hand over comes to you with the thread attached.
You write the questions and the criteria. The agent asks the questions in the order you set, on every channel, and records the answers in the thread. The criteria for offering a consultation are in its instructions too: a minimum budget band, the project types you take, the regions you work in. An enquiry that meets them is offered a slot; one that does not gets the reply you wrote, which might be a referral to a smaller practice or a polite decline. The agent does not judge the project; it applies your rules and keeps the answers so that the architect who takes the meeting has read them before the client arrives.
From your own pages and documents, yes. The agent reads and stores the pages of your site you give it and links to them verbatim when a prospect asks which of your projects is like theirs or how the stages run. Add your practice profile as a document and your fee structure and lead times as data points if you want them quoted. It answers only from that material and links only to pages you allowed. A question it was not given the answer to, a fee for a specific project, a timeline for a specific site, goes to your inbox with the agent standing down, because the practice decides what may be said.
It books. The agent is given your real open slots for the week ahead from Google Calendar, Microsoft 365, Calendly, Cal.com or one of the other supported calendars, offers a few of them, and when the client picks one, checks it against the slots that were offered, writes it into the calendar and confirms it in the reply. It cannot book a time that was not free and cannot cancel an appointment it was not shown. If the practice has two directors who each take their own first meetings, run one agent per director and each books into its own calendar.
It sends what the sequence says and no more. You write the steps and the gaps between them, a note at day 7, a completed project at day 30, a check-in at day 60, and each step can be sent only to people who opened, or did not open, the one before. Every email goes from your own mailbox inside the hours you set, carries a one-click unsubscribe, and the sequence ends on every channel the moment the person replies, books or opts out. One reply per thread, a daily cap per sender and hourly caps per mailbox mean it cannot overshoot. It is more disciplined than a person with a reminder in their diary, not less.
Yes. Connect Instagram, and Messenger, WhatsApp, Telegram, LINE, Viber, Slack, Discord, Bluesky, Reddit and X if you use them, with your own app credentials, and the same agent answers there from the same pages with the same qualifying questions, inside each platform's reply window. A DM asking "do you do extensions like this?" is answered from your project page and asked where the site is. The DM lands in the same inbox as the calls and emails, tagged with the platform, and because every handle a person uses is attached to one record, the prospect who DMs in January and emails in March is one person with one history.
Only if you gave it the figure. The agent quotes fee structures, minimums and typical lead times that you loaded as data points, and nothing else. A prospect asking what their specific project would cost, or when it could be on site, gets the reply that an architect will discuss that at the consultation, and the thread comes to your inbox with the agent standing down. Every reply is stored beside the material it was generated from, so you can read back exactly what any prospect was told about money or time and where it came from. The rule is in its instructions, and the practice decides what may be said.
The same as for a large one: $25 a month or $200 a year for the whole practice, with unlimited agents, contacts and lists, and a 7-day free trial. There is no per-seat, per-call, per-conversation or per-contact charge. The AI runs on your own key with one of sixteen providers, billed to you by that provider at cost, typically a few dollars a month for an agent handling a practice's enquiries. Calls and texts are billed by your carrier at its rates. A practice of three pays what a practice of thirty pays.
One subscription, unlimited agents. What changes is the knowledge base, the channels and the instructions.
Build the agent, connect the site, the line and the calendar, switch inbound on. Every enquiry is answered and qualified, every consultation is booked, and every prospect is kept in touch with until they decide.
7-day free trial · Cancel anytime