Stop answering the same questions twice
AI Chatbot for
Your Website.
Upload documentation. Kodda indexes it, chunks it, and serves grounded answers via a chat widget on your site. No-code setup, RAG-powered responses in under 2 seconds.
Start free → Deploy when ready → Scale with widget, API, and more
Why teams choose Kodda
Private by default
Your documents stay scoped to your workspace. Not used for model training.
Answers from your docs
Every response traces back to source material you uploaded.
Deploy fast
Public link, embedded widget, or API — one bot, three channels.
How It Works
Docs in. Grounded answers out.
Seven-stage pipeline: upload, extract, chunk, embed, store, retrieve, generate. Each stage is deterministic and auditable.
UPLOAD
Add documents
EXTRACT
Parse text
CHUNK
Split by topic
EMBED
Generate vectors
STORE
Index in Qdrant
RETRIEVE
Semantic search
GENERATE
RAG response
7
Pipeline Stages
Fast
Vector Embedding
100%
Workspace-Isolated
See the full Kodda workflow in motion
A short product reel that shows setup, knowledge ingestion, and grounded chat responses without forcing another live demo into the hero itself.
Core Workflow
What you can deploy with Kodda
Upload docs. Kodda indexes, chunks, and embeds them. Deploy a grounded assistant via public link, widget, or API.
Auto-answer repeat questions
Point your FAQs, policies, and service docs at Kodda. Grounded responses replace manual replies.
Embed on your site. No dev required.
Publish a public page or drop the chat widget into your site with a single snippet.
Workspace-scoped. Not training data.
Documents stay in your workspace. No public model training. No data leakage.
Source citations on every answer
Every response traces back to the uploaded source. Your team can verify what the assistant says.
Import from the tools you already use
Upload files, connect Google Drive or Notion. Kodda handles extraction, chunking, and embedding.
Escalate to a human when needed
When the assistant should not guess, visitors get a clear path to reach your team.
Multi-language responses
Automatically detect and respond in the language your visitors use — no manual configuration required.
Supported File Types
No AI team required
Upload docs. Configure indexing. Deploy a grounded assistant. Kodda handles extraction, chunking, embedding, and retrieval — you ship the knowledge.
6+
File types parsed
Scoped
Workspace-isolated data
Sourced
Every answer cited
24/7
Uptime on deployed bots
Index existing docs
FAQs, policies, service pages, notes — upload and go
Private, grounded responses
Scoped to your workspace. Every answer traces to source
Deploy without engineering
Public link, embedded widget, or API endpoint
Pain Points
Every unanswered question is time your team won't get back.
Visitors wait, email, or leave. Your team fields the same questions again. The answers already exist in your docs.
Index docs you already have
Private by default
Deploy to your site in minutes
Customers wait on questions your docs already answer
Service details, pricing, policies, onboarding — the answers exist, but they're not where visitors ask.
Knowledge scattered across files, folders, and notes
The right answer lives in docs, PDFs, Drive folders, and old messages — not in one queryable surface.
Responses vary by who replies
Manual answers drift. No consistency, no audit trail, no single source of truth.
Teams trapped in repetitive reply loops
Founders and support staff answer the same questions instead of shipping product.
Kodda indexes your docs and deploys a grounded assistant.
Visitors query your knowledge base directly. Grounded answers, sourced citations, no dependency on team availability.
Plans
Transparent pricing. No per-message surcharges.
Start free. Upgrade when you need more bots, storage, or channels. Every paid plan includes full API access and white-labeling.
Free
Evaluate with your own docs. No credit card.
- 1 Bot
- 1 Library
- 15MB Uploaded File Storage
- 100 messages / month
- Public bot link
- Source citations
- Website embed support
- API key access
Starter
Full API access. White-label widget. Embed on your site. No branding tax.
Save 40%
- 2 Bots
- 2 Libraries
- 20MB Uploaded File Storage
- 1500 messages / month
- 7-day free trial included
- Website embed support (iframe + embed.js)
- API key access
- Platform integrations (Slack, Discord, Zapier, Webhooks)
Pro
Full API access. White-label widget. Embed on your site. No branding tax.
Save 40%
- 3 Bots
- 3 Libraries
- 50MB Uploaded File Storage
- 5000 messages / month
- 7-day free trial included
- Website embed support (iframe + embed.js)
- API key access
- Platform integrations (Slack, Discord, Zapier, Webhooks)
More plans coming soon
We're starting with a small public beta. As we learn from real usage and operating costs, we'll add better-fit plans and pricing options.
Pricing will evolve as beta usage and operating costs become clearer.
- Team collaboration
- Custom branding
- Scheduled reports
- Built to evolve with real customer usage
- New plans will ship as the economics make sense
- Current pricing stays intentionally simple during beta
More plans coming as we learn from real usage.
Usage Add-ons
Extend your monthly capacity on-demand without changing your base subscription.
+5,000 messages / month
Extend monthly capacity on-demand.
$19/mo
Checkout is handled by RevenueCat. Final taxes and currency conversion are shown at checkout.
Compare Plans
| Feature | Free | Starter | Pro |
|---|---|---|---|
| Bots | 1 | 2 | 3 |
| Libraries | 1 | 2 | 3 |
| Storage | 15MB | 20MB | 50MB |
| Messages / month | 100 | 1,500 | 5,000 |
| Website embed | |||
| API key access | |||
| Platform integrations | |||
| Auto-sync | |||
| Advanced analytics | |||
| Remove Powered by | |||
| Custom rate limiting | |||
| 7-day free trial |
Need help deploying your
first assistant?
Tell us what docs you have, which questions repeat, and where the assistant should run. We'll point you to the fastest setup.