work / AI / lead capture chatbot

Visitors get answers. Sales gets leads.

Website questions and leads were handled by hand. I built a chatbot that answers common questions, qualifies leads and sends them straight to the CRM.

role
Solo developer
client
Philippine automation agency
status
Delivered

Tools & platforms

DialogflowZapierHubSpotWix

AI & LLM

in the product
Dialogflow (intent matching)
to build it
ChatGPT

Skills

Conversational designCRM integrationLead routing
01// how I approached it

The thinking behind it

  1. 01 · What I noticed

    Visitors asked the same questions, and leads waited in an inbox.

  2. 02 · What I decided

    Answer the common questions instantly and send every lead straight into the CRM with a deal attached.

  3. 03 · What I delivered

    Leads and deals created automatically, with a human handoff.

02// old vs new

Old system vs new system

Old systempeople as the glue
  1. Visitor emails a question
  2. Someone replies later
  3. Lead copied into the CRM
  4. Sales maybe follows up
New systemthe system does the work
  1. Visitor chats any time
  2. Bot answers or qualifies
  3. Lead and deal created in the CRM
  4. Team alerted
03// how it works

Watch it run

Click through the four steps. In step 3, click any box to see what it does.

synthetic data · no client screens
Reconstruction with synthetic data · no client screens
04// the hard part

Answering instantly without losing the human touch.

Visitors wanted quick answers and sales wanted every lead. The bot answers common questions on its own, qualifies leads, creates the contact and deal in the CRM, alerts the team, and hands over to a person when needed.

05// key decisions

Why it's built this way

decision 1

Leads go straight to the CRM

Every chat that matters becomes a contact and a deal.

Trade-off: CRM fields must be mapped carefully.
decision 2

Human handoff

The bot knows when to pass the conversation on.

Trade-off: someone must be ready to respond.
decision 3

Team alerts

Sales hears about new leads immediately.

Trade-off: alerts need tuning to avoid noise.