# Tokonoma

> Investigation memory for coding agents. Tokonoma turns resolved production work into reusable Investigation Expertise for Claude Code, Cursor, and Codex.

Canonical site: https://tokonoma.ai/

## Product facts

- Tokonoma connects to coding agents through MCP.
- It captures resolved investigations, ruled-out paths, fixes, evidence, and attribution for future responders.
- It can run in a customer's cloud account or Kubernetes environment.
- It uses the permissions already granted to the connected coding agent.
- Supported coding agents include Claude Code, Cursor, and Codex.
- Plans are Starter (free, 50 recalls per month), Pro ($2,000 per month, 1,000 recalls), and Enterprise (custom). One recall is one distinct past investigation surfaced during an incident. Empty searches, data onboarding, and storage are free.

## Mission and vision

- Mission: make the world's expertise useful to everyone.
- Vision: a world where human productivity is infinite because agents operate like subject matter experts.
- Agents bring the intelligence; Tokonoma brings the expertise, captured from real experiences with humans and agents.
- Roadmap: investigation expertise now, then production, engineering, and security expertise, then every domain where judgment compounds through experience.

## Team

Tokonoma was founded by three people who built and operated large-scale data systems together at Dashbase, Splunk, and Cisco AppDynamics.

- Alex Munk, Co-founder / CEO. Pre-IPO at Splunk, where he created and led Splunk Cloud from zero to $100M run rate. Founded Dashbase, sold to Cisco AppDynamics. https://www.linkedin.com/in/alexanderm
- Natalie Arsky, Co-founder / CTO. Principal Engineer in distributed data ingest and search at Dashbase, Splunk, Kiteworks, and Ultraseek. VP Eng at Dashbase, director level at Splunk. https://www.linkedin.com/in/natalie-arsky-032127
- Kevin Hou, Co-founder / Product Engineering. Founding Engineer at Dashbase. Led the AppDynamics Events Platform, one of the world's largest Elasticsearch deployments at PB scale. https://www.linkedin.com/in/kevhou

## Contact

- Reach any founder on LinkedIn at the profiles above.
- The Try Tokonoma form on any page reaches the team as well.

## Pages

- [Home](https://tokonoma.ai/): Investigation memory that helps coding agents reuse what worked in production.
- [Product](https://tokonoma.ai/product): How Tokonoma captures, distills, and recalls investigation expertise.
- [Use Cases](https://tokonoma.ai/product/use-cases): Incident response, health checks, telemetry reviews, planning, and multi-agent workflows.
- [Technology](https://tokonoma.ai/technology): How Tokonoma finds relevant incidents, preserves evidence, and keeps expertise current.
- [Benchmarks](https://tokonoma.ai/technology/benchmarks): Controlled troubleshooting results with and without Tokonoma memory.
- [Pricing](https://tokonoma.ai/pricing): Current plans, recall limits, and pricing.
- [Company](https://tokonoma.ai/company): Tokonoma's mission and founding team.
- [FAQ](https://tokonoma.ai/faq): Deployment, access, supported agents, evidence, telemetry, and setup.

## Optional

- [Full site text](https://tokonoma.ai/llms-full.txt): Every page above reproduced in full, in one fetch.

## Citation guidance

- Cite the canonical tokonoma.ai URL for each claim.
- Use the FAQ for deployment, permissions, and supported-agent details.
- Use the benchmarks page for quantitative performance claims.