Pricing Plans
Choose the engagement model that best fits your app build and rollout
Agent Foundation Sprint
A fast, structured sprint to ship a scoped v1 agent with the right tools, guardrails, and a clear path to production.
- Defined scope with clear deliverables
- 1 core workflow (v1) + success criteria
- Tool plan (MCP / actions) + least-privilege design
- Initial evals/graders to prevent drift
- Deployment recommendation (Agent Builder vs SDK runtime)
- Handoff docs and next-steps plan
Production Build
Productionize an agent end-to-end: hardened integrations, approvals, eval gates, and rollout strategy.
- MCP/tool enablement (schemas, auth, rate limits)
- Approvals + permissions for sensitive actions
- Evals factory: datasets + graders aligned to KPIs
- Agent Builder publish OR code-first SDK runtime
- Version pinning + rollout strategy + regression suite
Agent Ops Retainer
Ongoing iteration, monitoring, and expansion as usage grows and new tools/integrations are added.
- Ongoing workflows and tool improvements
- Regular eval refresh and drift prevention
- Reliability monitoring and performance tuning
- Permission/scoping reviews as usage grows
- Priority support and fast iteration
Need a Custom Solution?
We understand that every business has unique needs. Contact us to discuss a customized plan that works for you.
Book a 30-min build consultFrequently Asked Questions
Project costs are determined based on scope, complexity, delivery needs, and resource requirements. We provide detailed estimates after understanding your specific needs during the initial consultation.
Agent Builder is great for multi-step workflows where you want shared visibility, rapid iteration, and trace-based evals. The Agents SDK is better when you need deep customization, custom runtime requirements, or code-first control from the start.
We typically deploy via ChatKit (embed a chat UI wired to your workflow) or export the workflow code to run and customize with the Agents SDK—based on your product and infrastructure needs.
Not currently. Agent Builder is best for text-based multi-step workflows; if voice is required, we’ll build it with an SDK-based runtime and the right safety/approval patterns.
Yes. We design and implement MCP servers that expose only the minimum tools and data needed, with clear permissions, auditability patterns, and environment separation.
We design for least privilege: scoped data access, explicit action permissions, and clear privacy boundaries. We also recommend audit logging and approvals for sensitive actions when appropriate.
If directory distribution is part of your goals, we can help with listing copy, permissions/data disclosure clarity, UI polish, and review-readiness (subject to platform requirements and approval).
Retainer agreements include ongoing development support, maintenance, regular updates, and priority support. The exact scope is customized based on your needs and usage requirements.
Yes, you can switch between plans as your needs evolve. We'll work with you to ensure a smooth transition and continuity of service.
We offer proof-of-concept phases for larger projects to demonstrate value and ensure alignment with your objectives before full implementation.