How long does it take to index my content?

Most sources are processed and indexed into Pinecone within minutes. You can track ingestion status as Synced, Pending, or Failed from your dashboard.

Can I customize how the chatbot looks and responds?

Yes. You can set the widget theme, brand colors, bot name, and avatar image, and choose between OpenAI and Anthropic models to control how the bot responds.

Can I see what my chatbot is being asked?

Yes. The conversations tab shows full chat histories, and the usage dashboard lets you monitor activity per workspace or filter down to a specific bot.

What happens if my workspace hits its usage limit?

Service pauses automatically once a workspace limit is reached, so you never receive an unexpected overage bill.

Custom AI Chatbots

Deploy custom AI assistants trained on your own data.

TealSub lets you spin up dedicated workspaces, invite your team, and build custom AI chatbots in minutes. Connect your documents, style the interface to match your brand, and track usage with total transparency.

The Setup Workflow

From zero to a fully customized assistant.

Phase 1

Set up your workspace & team

Create your workspace to get started. Billing is scoped entirely to the workspace, not the account, giving you clean boundaries. Invite team members to collaborate, review pricing and limits beforehand, and manage everything from a single dashboard.

Workspace-isolated billing setups
Invite team members to collaborate instantly
Review usage limits directly from the usage page

Phase 2

Ingest your knowledge base

Upload PDFs, DOCX, Markdown files, or provide sitemaps and direct web URLs (avoid redirect links). The platform processes and indexes your files into Pinecone via high-performance embeddings within minutes, keeping track of status updates seamlessly.

Supports PDFs, DOCX, MD, sitemaps, and direct URLs
Real-time ingestion status: Synced, Pending, or Failed
Easily upload or delete content sources anytime

Phase 3

Customize look, feel & voice

Make the chatbot truly yours. Head over to the bot settings to customize the layout theme, brand colors, bot name, and upload a custom avatar image. Tune OpenAI and Anthropic models to respond exactly how your brand sounds.

Tailor widget themes, colors, and branding
Upload custom bot names and avatar images
Powered by OpenAI and Anthropic models via RAG

Phase 4

Monitor conversations & track usage

Keep complete control post-launch. Review exact chat histories inside the conversations tab to see how your bot performs. Monitor usage graphs per workspace, or filter down to inspect granular consumption data for a specific bot.

Track real-time interactions in the conversations tab
Filter usage analytics down to the specific bot level
Flexible monthly or yearly cycles; switch or delete anytime

Infrastructure

Engineered for accurate retrieval and total control.

Frontend Architecture

Next.js, React, TypeScript, Tailwind CSS, and shadcn/ui power a lightning-fast, highly responsive user interface.

AI & Vector Search

OpenAI and Anthropic models drive contextual answer generation, while Pinecone acts as our high-performance vector database for fast RAG retrieval.

Data & Background Jobs

Supabase, Postgres, Drizzle ORM, and Trigger.dev manage secure data structures, instant syncing states, and automated workspace workflows.

Reliability & Limits

Upstash Redis handles strict rate limiting. To prevent surprise overages, services pause safely when workspace limits are hit.

Setup FAQ

Questions about getting started

Tealsub supports PDFs, DOCX files, Markdown files, sitemaps, and direct web URLs. Redirect links are not supported — use the direct URL of the page.

Still unsure? Ask your own bot a question — that's the whole point.

Get Started

Ready to spin up your workspace?

Set up your team boundaries, configure your pricing preferences, and index your first data source. Service stops automatically if limits are reached, ensuring no unexpected billing surprises.