Category Brief
Support is moving into the product
The next generation of customer support does not live in a help center or a ticket queue. It lives inside the product, at the exact moment a user gets stuck.
Market thesis
The problem is structural
Users encounter issues inside products — but current support tools live outside products. This structural mismatch means context is always lost in the handoff. Users describe symptoms; support teams diagnose from scratch. Every repeatable issue costs the same as a novel one.
AI changes what is possible
AI agents can understand product context, reason about user state, and guide through resolution steps. But only if they are embedded in the product with access to the right information. A chat widget with documentation is not enough.
Control is the bottleneck
Teams want AI support that does useful work, but they cannot trust systems without approval boundaries, trace logs, and operator visibility. The companies that solve the control problem will win the category.
Open source will win
Developers and product teams want transparency in systems that interact with their users. Open-source SDKs let teams audit, modify, and self-host. Proprietary black boxes will lose trust over time.
Competitive landscape
Traditional ticketing (Zendesk, Freshdesk)
Ticket-first model. AI extends an existing queue workflow rather than preventing tickets.
Chat widgets (Intercom, Drift)
Widget-based. Limited product context. No action execution capability.
Help centers (Notion, Guru)
Article-based self-service. User must search and self-diagnose. No proactive assistance.
ResolveKit
Embedded SDK with product context, approval system, trace logging, and open-source transparency.
Product vision
ResolveKit envisions a world where support issues are resolved at the source — inside the product, in the exact moment of need. The SDK embeds a product-aware AI agent that understands what the user is seeing, can explain what went wrong, and can execute approved actions to fix it.
Resolve upstream
Handle issues before they become tickets.
Control by design
Approvals, policies, and traces built in from day one.
Open and auditable
AGPL-3.0 licensed. Inspect, modify, self-host.
The economics of in-app resolution
The cost model for in-app support is fundamentally different from traditional support tools. When issues are resolved inside the product, the economics shift:
Lower per-resolution cost
AI-assisted resolution inside the product costs significantly less than human-handled tickets. Self-hosted mode reduces costs further by eliminating per-resolution SaaS fees entirely.
Compound improvement
Every resolved case generates trace data that improves future handling. The system gets smarter and cheaper over time as more patterns are recognized and automated.