AI Agents and Assistants
What is an AI agent? A plain-English guide for business owners
What an AI agent is, how it works, where it helps a small business, the risks, and a 7-point check to see if you are ready for one.
AI Features & Bots
AI Website Chatbot for SaaS
Trial users and prospects get answers from your docs and pricing pages, and sales-ready questions reach a person.
Trial users who cannot find an answer quietly drop off before they ever talk to sales. This chatbot answers setup, how-to, feature and plan questions on your marketing site or inside your web app, using your docs, changelog and pricing pages. Account-specific and upgrade questions go straight to your team in Intercom, Zendesk or email. The code lives in your repository.
- Trial users get setup and how-to answers inside the product
- Upgrade-ready questions reach sales with the conversation attached
- Ships as code in your repository, so you own it
$3,500
20–30 working days
Optional maintenance $249/mo
What it is
AI Website Chatbot for SaaS is an assistant for your marketing site or web app that answers trial, setup, how-to and plan questions from your docs, changelog and pricing pages, and hands upgrade or account-specific questions to your team. It includes the SaaS workflow template and ships as code in your repository. It answers from your current docs.
Why businesses buy it
SaaS teams lose trials in the quiet moments: a user stuck on setup at night, or a prospect unsure which plan fits. Support tickets pile up with questions your docs already answer. This chatbot answers from those docs instantly and passes buying signals to a person, so more trials reach their first win and sales sees warm leads sooner. The written scope fixes the content, intents, handoff and testing before you order, so there are no open-ended surprises.
Problems this solves
Trial users get stuck on setup and quietly drift away
The chatbot answers setup and how-to questions from your docs and changelog, on your marketing site or inside your web app. Trial users get a direct answer rather than waiting on a support ticket.
Pricing page visitors cannot tell which plan suits them
Plan questions are answered from your pricing pages using the SaaS intent library. Upgrade and account-specific questions go to your team, so nothing about a customer's own account is guessed.
Support answers fall out of step after each product release
Answers come from your docs and changelog rather than a fixed script, so they follow the content you keep current. Forty documented test questions with pass/fail results check accuracy at launch.
A good fit if
Trial users ask the same setup and feature questions your docs already answer, and they ask before they ever open a ticket.
Prospects on the pricing page want plan differences explained, and a slow reply costs you the signup.
You want the assistant inside your web app as well as on the marketing site, delivered as code your engineers can review.
Questions about upgrades, seats or a specific account should go to sales or support in Intercom, Zendesk or by email.
Not the right fit if
You need it to read live billing data and make plan changes in Stripe. See AI Customer Support Agent for SaaS, which takes approved account actions.
You want new customers guided through setup milestones with nudges when they stall. See AI Customer Onboarding Agent for SaaS.
You want trial signups qualified and scored for sales. See AI Lead Qualification Agent for SaaS.
You need it in a second language or on a second channel such as Slack. Add the Additional Language Pack or a channel add-on at checkout.
How it works
- It learns from your product content. Up to 100 documents or 200 web pages, typically help docs, changelog, pricing and security pages, all approved by your team.
- Five SaaS intents are set up. Usually trial questions, setup and onboarding, how-to and troubleshooting, plans and pricing, and upgrade or renewal requests.
- It sits where users are. A branded widget on your marketing site or inside your web app, shipped as production code to your repository with no secrets in the code.
- Account-specific questions are handed over. Seat changes, invoices or a stuck integration go to your Intercom or Zendesk inbox, or to email, with the conversation attached.
- It refuses to invent features. If the docs do not cover something, it says so and offers the handoff rather than promising a capability you do not have.
- It is proven on 40 test questions. Drawn from your real trial and support questions, each marked pass or fail before launch.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- Your help docs, changelog, pricing and security pages, as files or URLs.
- Your top 20 trial and support questions, with the answers you approve.
- Access to Intercom, Zendesk or the email address that receives handoffs.
- Repository or hosting access, plus a technical contact to review the widget code.
- Your logo and brand colors, and where the widget should appear on the site or in the app.
- SaaS-specific rules, for example which plan questions sales should always take, and admin access and API keys for the handoff system, never by email.
- One decision-maker who approves answers and the go-live.
- An API key for your own AI model account (for example OpenAI or Google Gemini), created for this build with a spending limit.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake verified: docs, top 20 questions, handoff inbox and repository access confirmed.
Weeks 1–2
Content loaded and the five SaaS intents built and tuned.
Week 3
Widget code merged to your repository; handoff to Intercom, Zendesk or email connected.
Weeks 3–4
40 test questions run; pass or fail results sent to you.
Week 4
Two rounds of changes to answers, intents or widget.
Go-live
Live on your site or app; recorded handoff; 30 days of support.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $3,500 |
| Express | 15–25 working days | $4,100 |
| Rush | 10–20 working days | $4,750 |
What you receive at handover
- The AI website chatbot for SaaS, deployed on your site or web app and shipped as code in your repository
- The SaaS intent library and workflow template, covering up to five intents plus human handoff
- A handoff connection to your support email or helpdesk
- A test report built on SaaS scenarios, with forty test questions and pass/fail results
- An admin guide and the recorded 60-minute handoff session
- 30 days of post-launch support
Expected benefits and how to measure them
- Can help trial users reach their first setup steps without opening a ticket
- Helps support keep its time for account-specific issues
- Makes plan and pricing answers consistent for every visitor
- How-to and setup tickets
- Tag these in your helpdesk. Compare weekly counts for four weeks before launch with four weeks after.
- Trial-to-paid conversion
- Take it from your billing or CRM reports. Compare trial cohorts that started before launch with cohorts that started after.
- Upgrade conversations handed to sales
- Count upgrade handoffs arriving in your team inbox each week. Follow them through to closed deals in your CRM.
Before you start: Export a month of helpdesk tickets by tag and note your current trial-to-paid conversion. List which docs and changelog pages the chatbot will use.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build itself is a one-time payment, shown before you buy. Running it costs roughly $30 to $150 a month for AI model usage and hosting, billed in your own accounts, plus whatever you already pay for Intercom or Zendesk.
Usage during build and testing is billed to you directly, so your cloud and AI provider accounts stay yours.
Ask about running costsCustom requirements
Answers from a private customer knowledge base behind login, a second product's docs beyond the 100-document limit, or a sixth intent are outside the Scope Card. Add a note at checkout for a written Change Quote. Larger or bespoke projects go through the Custom AI request form instead.
Request a custom versionKeep it running
We recommend monthly maintenance for this build. It covers monitoring, fixes and monthly hours to keep answers current with each release, refresh pricing details and tune intents. Bigger changes for a major launch go through a Change Quote. Maintenance starts after the included support period and can be cancelled any month.
$249/mo (the minimum fee, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | A chat widget vendor bills per seat or per resolved conversation | An agency quotes after discovery calls | Ours is one fixed price, shown before you buy |
| Code ownership | A hosted widget stays in the vendor's platform | Depends on the contract | Ours is production code delivered to your repository |
| Feature answers | A generic bot may describe features you do not have | An agency build depends on what was agreed | Ours answers only from your docs and says when it cannot |
| Delivery | A widget installs fast but your content still needs tuning | An agency usually takes several weeks, set per project | Ours has a fixed delivery time shown before you buy, with faster options |
Questions about this product
Can AI Website Chatbot for SaaS run inside our web app, not just the marketing site?
Yes. The channel is one website widget, and it can be embedded in your logged-in web app or on your marketing site. The code is delivered to your repository, configured through environment settings with no secrets in the code, so your engineers review and deploy it like any other change.
Will it answer questions about a customer's own plan or invoice?
No. It has no access to billing data, so account-specific questions such as seat counts or a particular invoice are handed to your team in Intercom, Zendesk or by email. For an assistant that reads Stripe and makes approved plan changes, see AI Customer Support Agent for SaaS.
How does it stay accurate after each product release?
It answers from the docs and pages you approved, so updating those sources is the first step. Monthly maintenance covers reloading content and adjusting intents after releases within its change hours. Without maintenance, your team can update the sources and ask us to refresh.
How is the SaaS edition different from the standard chatbot?
The SaaS edition adds the SaaS workflow template, intents tuned for trials, plans and onboarding, and testing on real SaaS questions such as setup steps and plan limits. It also ships as code in your repository. Both prices, and the delivery speeds for each, are shown on their product pages before you order.
Which five intents should a SaaS company choose?
Most choose trial questions, setup and onboarding, how-to and troubleshooting, plans and pricing, and upgrade or renewal requests. You can swap one, for example security and compliance questions for an enterprise-focused product. Anything outside the five intents still gets the human handoff rather than a guess.
Can it hand a hot upgrade lead straight to sales?
Yes. Upgrade and expansion questions can be one of your five intents, with the handoff set to the sales inbox or a Zendesk or Intercom team. The conversation goes across with it, so the salesperson sees what the user asked. Lead scoring is not included.
Guides on this topic
All guides Answer trial questions before the trial ends.
Order AI Website Chatbot for SaaS and your site or app starts answering trial, setup and plan questions from your docs, with account and upgrade questions passed to your team.
Works well with
Front Line
AI Customer Support Agent
Resolves customer questions across chat and email using your knowledge base and hands off complex cases.
Lifecycle
AI Customer Onboarding Agent
Guides new customers through setup milestones and nudges when they stall.
Health & Risk
AI Churn Prevention
Predicts churn risk and recommends the retention play for each at-risk account.