WS Form stores forms in wsf_form (form structure as JSON), submissions in wsf_submit, submission data in wsf_submit_meta, and activity in wsf_log. The form definition includes every field, its type, its label, conditional logic, and validation rules. Submissions store one row per entry plus key-value meta for each completed field.

SleekAI's data sources map both definition and submission data into named variables. The visitor-facing bot knows which fields the current form expects, which are required, and which validations apply, so it can answer "what does field X mean" or "can I leave that blank" without staff. The admin-only bot can summarise this week's submissions, surface common patterns, and answer questions like "how many enterprise leads did the contact form get in the last 7 days" in one message.

Display conditions scope each bot by post type, taxonomy term, user role, logged-in state, and URL pattern. A long form embedded on a landing page gets its own bot, a returning customer in the account area gets a different one, and an editor browsing the dashboard gets a third. Every chat is logged inside WordPress with the model name, token usage, and origin page.

Forms are where visitors leave. The longer the form and the more important the fields, the higher the abandonment rate. Visitors hit a field that confuses them, hesitate, and close the tab.

The standard questions are predictable: is this required, what does this field mean, can I come back later, why did this new field appear when I picked the previous option. A generic chatbot makes that worse, because it does not know the form. It invents required-field rules, contradicts the validation messages, and quotes save-and-resume behavior that does not exist on this site.

The result is more support tickets, more abandoned forms, and lost leads. WS Form already stores everything the bot needs. The form definition with its fields, types, validation, and conditional logic sits in wsf_form.

The submissions sit in wsf_submit and wsf_submit_meta. SleekAI maps both into the chatbot prompt with point and click data sources, so the chat agrees with the form on every detail. Per-bot display conditions then let you run a visitor-facing form helper and an admin-only submissions analyst from the same install, each scoped exactly where it earns its keep and each running on a key the site owner controls.