AI chatbot development

    AI chatbot development: bots that answer from your content and know when to hand off.

    We build AI assistants grounded in your own documents, policies and systems, with guardrails, human handoff and measurable accuracy - for your website, WhatsApp, SMS or internal teams.

    Free 30-minute consultation · Reply within one business day

    At a glance

    Starting price
    From $4k
    Typical first release
    2–4 weeks
    Pricing model
    Fixed price per scope
    You own
    Code, data & accounts
    • Free 30-minute consultation
    • Written scope before any work
    • Working demo every week
    • 30 days of support after launch

    A good business chatbot does three things well: answers questions correctly from your information, captures and qualifies the people who are ready to buy, and hands the conversation to a human the moment it's out of its depth. Everything else is decoration.

    We build chatbots with retrieval-augmented generation (RAG) - the model answers from your approved content rather than guessing - and we test them against a set of real questions before launch, so you know the accuracy before your customers do.

    What we build

    AI Chatbots: what it looks like in practice.

    Website support assistants

    Answer product, pricing and policy questions 24/7 from your help center and documents, with links to the source.

    Lead qualification bots

    Ask the right questions, score the lead, book the appointment and push everything into your CRM.

    WhatsApp & SMS assistants

    The same assistant on the channels your customers already use - see WhatsApp automation.

    Internal knowledge assistants

    Staff ask questions of SOPs, contracts and manuals instead of interrupting the one person who knows.

    Booking & order status

    Connected to your scheduling or order systems so the bot can actually check availability or status, not just talk about it.

    Human handoff & inbox

    Seamless escalation to your team with the full conversation, in a shared inbox or your helpdesk.

    Pricing

    What it costs, in real numbers.

    Start with a pilot on one channel and one knowledge base, measure it, then expand. Typical ranges:

    Pilot

    From $4k

    One channel, one knowledge base, human handoff, a tested set of real questions and an accuracy report.

    2–4 weeks

    Production assistant

    $12k–$35k

    Multiple channels, CRM or helpdesk integration, lead capture, analytics and admin tools to update content.

    6–10 weeks

    Agentic workflows

    $35k+

    The assistant takes actions in your systems - bookings, refunds, order changes - with approvals and audit logs.

    3+ months

    AI model usage is billed per use by the provider (OpenAI, Anthropic and others), plus hosting and any messaging-channel fees. For most small-business volumes that's tens to a few hundred dollars a month; we estimate it from your real traffic before you commit.

    Get your fixed-price quoteWritten scope and a fixed number before any work starts.
    How it works

    From first call to launch, no surprises.

    A written scope and a fixed price before any work starts, a working demo every week, and a clean handoff at the end.

    1. Step 011 week

      Use-case & content audit

      Which questions matter, where the answers live, and which ones must always go to a human.

    2. Step 02Part of week 1–2

      Evaluation set

      We collect 50–200 real questions with correct answers - the yardstick for accuracy.

    3. Step 031–3 weeks

      Build the pilot

      Retrieval over your content, conversation design, guardrails and handoff.

    4. Step 041 week

      Test & tune

      Run the evaluation set, fix wrong answers at the source, and report accuracy before launch.

    5. Step 05Ongoing

      Launch & monitor

      Go live on one channel, review conversations weekly, then expand.

    Stack & integrations

    Proven technology, your accounts.

    Technology we use

    Models
    OpenAI APIAnthropic Claude APIOpen-weight models when data must stay private
    Retrieval
    PostgreSQL + pgvectorDocument parsingHybrid search
    Channels
    Website widgetWhatsApp Business PlatformSMS (Twilio)Slack / Teams
    Safety
    GuardrailsPII redactionHuman handoffConversation logs

    Systems we integrate with

    Named platforms are examples we build against through their public APIs - not partnership claims.

    HubSpotSalesforceZendeskIntercomGoogle CalendarCalendlyShopifyWhatsApp Business PlatformTwilio
    Is this right for you?

    An honest fit check.

    A strong fit if…

    • Your team answers the same 30 questions every day.
    • Leads arrive after hours and go cold before anyone replies.
    • You have written content - help docs, policies, price lists - to ground the answers.

    Probably not the right move if…

    • You need a quick FAQ widget - an off-the-shelf tool like Intercom Fin or Chatbase may be cheaper; our [AI chatbot cost guide](/guides/ai-chatbot-cost) compares them.
    • Answers depend on judgment calls that must never be automated.
    Expert notes

    What we'd tell you on the first call.

    Why accuracy is measured, not promised

    Every chatbot sounds smart in a demo. The difference is how it handles the awkward questions. We build an evaluation set from your real inbox and chat logs, run it before launch and after every change, and share the results. Wrong answers get fixed at the source - usually a missing or contradictory document.

    Custom build or subscription tool?

    Subscription chatbots charge per seat or per resolution and work well for standard help-center Q&A. Custom wins when the bot must act inside your systems, follow your compliance rules, run on WhatsApp, or when per-resolution fees at your volume exceed the cost of owning it. See our custom AI chatbot vs. off-the-shelf breakdown.

    Proof

    Software our team built and runs.

    We don't just build for clients - we operate our own SaaS products in production, so we design for the unglamorous parts: billing, support, uptime and change.

    Billing & collections · SaaS

    Wasooli

    Recurring billing and collections with WhatsApp delivery, digital payments, automatic ledgers, attendance and a customer app.

    • Automated monthly billing
    • Digital payment collection
    • Self-updating ledgers
    • WhatsApp team inbox + AI
    FAQ

    AI Chatbots: questions buyers ask.

    How much does a custom AI chatbot cost?

    Our pilots start around $4k for one channel and one knowledge base. Production assistants with integrations typically run $12k–$35k, plus AI usage fees that depend on volume.

    Will the chatbot make things up?

    We ground answers in your approved content, instruct the model to say when it doesn't know, route sensitive topics to humans, and measure accuracy against real questions before launch.

    Can it work on WhatsApp?

    Yes, through the official WhatsApp Business Platform. WhatsApp's policies allow businesses to use AI for customer service; general-purpose AI assistants are not permitted, so we design for your business use case.

    Can the chatbot book appointments or update our CRM?

    Yes. We connect it to your calendar, CRM or order system so it can check availability, book, and log leads with the conversation attached.

    Which AI model do you use?

    We choose per project - commonly OpenAI or Anthropic models - based on accuracy on your evaluation set, cost and data requirements.

    Is our data used to train the AI model?

    We use API providers and settings where your business data isn't used to train their models, and we document where data is stored.

    Work with us

    Got a brief? Let's build it.

    Thirty minutes, no pitch deck. We will tell you what we would build, what it costs, and whether we are the right team for it.

    No obligation · We reply within one business day