Amplify Your Product
Ship AI-Native UX in Days
Pricing aligned to growth - not overhead
Starter Copilot
For SaaS teams adding AI guidance and in-product assistance.
Features you'll love:
- ✔ Knowledge & Smart Q&A
- ✔ Intent Navigation
- ✔ Screen Awareness (Share Screen)
- ✔ Actions (simple, in-product workflows)
- ✔ Support Handoff
KINETIC
For teams expanding into fully agentic experiences with advanced intelligence.
Everything in Starter Copilot, plus:
- ✔ Task Agents (purpose-built AI operators)
- ✔ Multi-Model Provider (best model per task)
- ✔ Voice (voice-to-action experiences)
- ✔ Analytics Insights
- ✔ Agentic Evals
Scale
Governance, observability, and cost-efficient AI at enterprise scale.
Everything in KINETIC, plus:
- ✔ Single Sign-On (SSO)
- ✔ Advanced PII Masking
- ✔ High availability (multi-providers)
- ✔ Progressive Model Deployment
- ✔ Multi-Product Support
Feature Comparison
| Feature | Starter | Kinetic | Scale |
|---|---|---|---|
|
Actions
i
Let the agent trigger in-product workflows — creating, updating, or viewing entities and running simple flows directly from the conversation.
|
✓ | ✓ | ✓ |
|
Knowledge & Smart Q&A
i
Connect product docs, help center, and internal content so users can ask natural-language questions and get precise, cited answers from the agent.
|
✓ | ✓ | ✓ |
|
Intent Navigation
i
The agent interprets user intent and navigates to the right view, page, or dashboard — no menus needed.
|
✓ | ✓ | ✓ |
|
Screen Awareness (Share Screen)
i
The agent understands what the user sees, enabling guidance, highlights, and step-by-step assistance during complex workflows.
|
✓ | ✓ | ✓ |
|
Support Handoff
i
Seamlessly escalate from the agent to human support, passing full context so users don’t repeat themselves.
|
✓ | ✓ | ✓ |
|
Task Agents
i
Specialized agents built for specific tasks — analysis, summaries, planning — each with tailored tools and reasoning.
|
— | ✓ | ✓ |
|
Tandem (Shared State)
i
Keeps the agent in sync with the product’s entities and on-screen context so it reasons about the same objects the user sees.
|
— | ✓ | ✓ |
|
Multi-Model Provider
i
Choose the best model per task automatically — balancing cost, quality, and latency — without changing your agent interactions.
|
— | ✓ | ✓ |
|
Voice to Action
i
Users speak naturally — the agent interprets, executes actions, navigates, or answers instantly.
|
— | ✓ | ✓ |
|
Analytics Insights
i
Understand which actions users take, what the agent completes, and measure value delivered inside your product.
|
— | ✓ | ✓ |
|
Agentic Evals
i
Evaluate agent behaviors using golden paths and automated tests to safely improve prompts and models.
|
— | ✓ | ✓ |
|
Single Sign-On (SSO)
i
Enterprise authentication for giving users secure, governed access to the agent using existing identity systems.
|
— | — | ✓ |
|
Advanced PII Masking
i
Automatically redact sensitive data before it's used by the agent or sent to models — ideal for enterprise compliance.
|
— | — | ✓ |
|
High Availability (Multi-Providers)
i
Keep the agent online by routing across multiple model providers when one becomes slow or unavailable.
|
— | — | ✓ |
|
Progressive Model Deployment
i
Roll out new models to a subset of users, compare performance, and adopt them safely when they outperform the current one.
|
— | — | ✓ |
|
Multi-Product Support
i
Deploy the same agent across multiple products or business units with centralized governance and configuration.
|
— | — | ✓ |
FAQ
Commonly Asked Questions
Yes. We have a Startup Program with discounted pricing for early-stage teams. If you’re under roughly 20 employees or at seed stage, you can speak with our sales team and we’ll shape a plan that fits your size and roadmap.
Starter – If you want to add an in-product AI agent quickly for guidance, navigation, and basic actions.
Kinetic – If you’re ready for full agentic workflows with Task Agents, multi-model support, voice, and analytics.
Scale – If you need governance, SSO, multi-product support, high availability, and tighter controls for larger teams.
If you’re unsure, share your top 2–3 use cases with our team and we’ll recommend a starting point.
An Action is any workflow the agent triggers inside your product: creating or updating an entity, pulling a specific view, navigating to a page, or running a defined task in the background.
No. Foldspace connects via a lightweight front-end SDK to the services you expose. You define the Actions; the agent calls them. Your existing architecture and release cadence stay intact.
Starter (Copilot): Most teams are live within 2–4 hours. This includes connecting your knowledge base, crawling content, and installing the Foldspace tag/SDK to enable navigation and screen awareness.
Kinetic: AI front-end developers and AI PMs usually create fully agentic experiences in 2–5 hours, depending on how many workflows and Task Agents they include.
Yes. You can use our Chrome extension to preview the agent on top of your existing UI without engineering work. For deeper evaluation, we support short, time-boxed POCs.
You can bring your own LLM providers or use our defaults. Kinetic and Scale support multi-model routing, so each task can use the most suitable model for quality, latency, and cost.
We’re designed to minimize data exposure. You control what context the agent can access, and we avoid storing customer data by default. The Scale plan adds advanced PII masking, SSO, auditability, and multi-provider routing for stricter environments.
Yes. With Scale, you can run a single agent across multiple modules or products, while keeping configuration, policies, and telemetry centralized.
No. Foldspace sits alongside your current stack. It adds agent-level telemetry (actions, completions, adoption) that you can combine with tools like Pendo, Amplitude, or in-house analytics.
Starter includes standard support and documentation. Kinetic and Scale add higher-touch onboarding, implementation guidance, and access to our product and AI team for deeper design and troubleshooting.
Yes. Scale is designed for higher compliance needs with advanced masking, SSO, governed access to Actions, and the ability to control which models and regions are used for inference.