White-label AI support end to end: reseller setup, custom domain, and your own email
Supportweave Team · July 25, 2026 · 9 min read

When you resell AI support, the product your clients log into should be *yours* — your name in the header, your logo on the login page, your domain in the address bar, and your name on every email it sends. That's the whole point of white-label: you own the relationship, you set the price, and the platform stays invisible.
This is the end-to-end setup guide. By the end you'll have a reseller plan running, a fully branded portal on your own domain, transactional email leaving your own provider account, and your first client workspace provisioned with credits and sensible caps. If you want the business case first, start with how agencies add a revenue line with white-label AI support and come back here to build it.
What you'll need
- A Supportweave account where you're the organization owner — branding, domain, and email settings are owner-only, so one client's admin can never rebrand your whole operation
- A subdomain you control, like
support.youragency.com - An account with an email provider (Resend, SendGrid, Mailgun, Postmark, Brevo, Mailjet, SparkPost, Amazon SES — or any SMTP server)
- About an hour, most of it waiting for DNS
Step 1: Subscribe to a reseller plan
Open Reseller in the dashboard sidebar, pick a tier, and click Start reselling — checkout runs through Stripe and your pooled credits are granted the moment the first invoice is paid.
Each tier is a flat monthly fee that buys three things: a pooled credit allowance your clients draw from, a number of client seats, and an org-wide knowledge-source allowance. Tiers scale from Accelerator up to Enterprise, which carries unlimited seats — the current tiers and prices are on the pricing page.
Two details worth knowing up front. Seats count resold clients only — your own workspace never consumes one, so a tier's seat count is a true count of clients besides yourself. And white-label features (branding, custom domain, your own email sender) are tied to holding an active reseller plan; you manage or cancel the subscription anytime through the billing portal, and if you ever let it lapse your saved branding and verified domain are kept, so re-subscribing restores everything instantly.
Step 2: Brand the portal
Go to Settings → Branding. This is org-wide — every client workspace under you inherits it.
- App name — replaces "Supportweave" across the dashboard and in email subjects and footers
- Logo — upload it right there (up to 1 MB); no need to host the file yourself
- Accent color — your brand color, applied across the portal chrome
- Support URL — where your clients' "help" links point (your site, not ours)
- "Powered by" badge — turn it off; nothing on the page says Supportweave
- Email from-name and reply-to — the name your clients see on every message, and where replies land
Once saved, the brand renders everywhere your clients look: the logged-in dashboard, the login and sign-up pages, and the transactional emails. One separate knob to remember: the chat widget your clients embed on *their* sites is branded per chatbot, on each bot's Appearance tab — so every end customer sees the right business, not your agency.
There's also a Reset action if you ever want to return the styling to defaults. It deliberately clears *styling only* — your custom domain and email delivery settings survive, so a cosmetic do-over can never break a live domain or your deliverability.
Step 3: Put it on your own domain
A branded portal on app.supportweave.com still isn't yours. Settings → Custom domain fixes that in three moves:
- 1Claim the hostname. Enter the bare domain —
support.youragency.com. The app issues a verification token and shows you exactly one DNS record to publish. - 2Prove you own it. Add the TXT record at
_sw-domain-verify.support.youragency.comwith the token as its value, wait for DNS to propagate (usually minutes), and click Verify. Only a verified domain ever resolves to your brand — a rival can't claim your hostname and hijack your look. - 3Point the traffic. CNAME the hostname at your deployment and let your hosting layer issue the certificate — Cloudflare's custom-hostname feature, Caddy's on-demand TLS, or your platform's custom-domain setting all work. HTTPS terminates at your proxy or host, so use whichever you already run.
From then on, everything pre-login on that host — the sign-in page your clients bookmark, the password-reset pages they land on — carries your brand. (Logged-in clients see your brand regardless of hostname, because the dashboard themes by organization.)
A few troubleshooting notes from the field:
- "Verification record not found yet" — TXT propagation can lag, especially with a high registrar TTL. Check with
dig TXT _sw-domain-verify.support.youragency.combefore clicking Verify again. - "That domain is already in use" — another organization holds a claim on it. Recently abandoned, never-verified claims are released automatically after a grace period; otherwise pick a different hostname.
- Apex domains (
youragency.comwith no subdomain) can't take a CNAME — DNS rules, not ours. Use your DNS provider's ALIAS/ANAME or CNAME-flattening record, or stick with a subdomain.
Step 4: Send email from your own provider
Out of the box, verification emails, team invites, lead notifications, and weekly digests are delivered by the platform sender, with your from-name and reply-to from Step 2. For full white-label — mail that authenticates against *your* domain from *your* provider account — take over delivery in Settings → Email delivery:
- 1Pick your provider preset — Resend, SendGrid, Mailgun, Postmark, Brevo, Mailjet, SparkPost, or Amazon SES — and paste that provider's credentials. The form pre-fills the host and username for you; there's a Custom SMTP option for anything else, with the standard submission ports to choose from. Credentials are encrypted at rest.
- 2Pass the test send. Saving marks the config pending; the app immediately sends a branded test email to you, the org owner, and activates your sender only when that send succeeds. No silent misconfiguration — if the test fails, the platform sender keeps delivering while you fix it.
- 3Set up SPF and DKIM for your sending domain at the provider (their docs are linked next to the form). This is the step that keeps your mail out of spam folders — don't skip it.
The failure mode is designed in your favor: if your reseller plan ever lapses, delivery falls back to the platform sender rather than dropping mail. Password resets and invites keep arriving no matter what.
Step 5: Provision client workspaces
Now the fun part. In Reseller → Client workspaces, enter a client's name, pick their starting plan, and click Provision client. You get an isolated workspace under your organization — its data fully separated, your brand already applied. Set up their first chatbot yourself or invite the client's team and let them build; either way an industry template gets a working knowledge base live in minutes.
When you're out of seats, the button tells you so and points at the next tier — nothing breaks, you just upgrade when the roster grows.
Step 6: Allocate credits and set the guardrails
Your clients' AI usage draws from your pooled credits, and each client row on the reseller page gives you the controls:
- Allocate credits — move credits from your pool to the client. This is what turns their metering on; their bots consume what you allocated, not the raw pool.
- Monthly spend cap — the ceiling one client can burn in a month. Set it on every client; it's what stops a runaway integration from eating the pool your other fourteen clients share.
- Per-client limits — override the plan defaults for chatbots, team seats, knowledge sources, and API access, in either direction: lift a starter client above their plan or hold a heavy one below it. Lowering a cap never deletes anything a client already built — it only blocks new creation until you raise it again.
If a client exhausts their allocation, their assistant pauses for visitors with a friendly message while human handoff keeps working — and you top them up (and probably upsell them) when they come asking. Your reseller panel shows the pool balance, seats used, and an org-wide knowledge-source meter, so you always know your headroom. When a client bumps into a limit, they see a neutral "contact your support provider" message — never a word about pools, tiers, or Supportweave.
What your client actually sees
Log in as a client once — it's clarifying. They get your name, your logo, your colors, on your domain, with your address on the emails. They build bots, add knowledge, embed the widget, and work their inbox. There is no billing page pitching them plans — billing shows "managed by your provider," because *you* are the provider. Their usage view shows the credits you allocated, not what those credits cost you.
That last part is the business model in one sentence: you charge each client what their support is worth — a retainer line, a monthly add-on, a usage package — while all of them draw from one flat-fee pool. A reseller with a full roster of clients each paying a healthy monthly rate is billing four figures against a single flat platform fee, and the per-client spend caps protect that margin automatically. Run your own numbers in the agency ROI calculator.
The whole setup, in order
- 1Reseller → pick a tier → Start reselling
- 2Settings → Branding → name, logo, color, from-name
- 3Settings → Custom domain → TXT verify → CNAME + TLS
- 4Settings → Email delivery → provider credentials → pass the test send → SPF/DKIM
- 5Reseller → Client workspaces → provision, allocate, cap
- 6Repeat step 5 every time you close a deal
The program details and packaging ideas live on the for-agencies page. If you'd rather walk through it live, book an agency demo and we'll set up your first branded client workspace with you.