# Jinba: enterprise workflows built and governed through chat

Canonical: https://mudpie.ai/companies/jinba/
Breadcrumb: [Home](https://mudpie.ai/) / [Companies](https://mudpie.ai/companies/) / [Jinba: enterprise workflows built and governed through chat](https://mudpie.ai/companies/jinba/)
Author: Ali Abouelatta (https://mudpie.ai/authors/ali-abouelatta/)
Published: 2026-09-19
Updated: 2026-09-19
Research type: Company profile
Method: Company and accelerator sources checked 2026-09-19. Product claims are attributed to their sources; this is research, not a hands-on product trial.

Jinba lets enterprise teams build, deploy, and use AI workflows through chat, flowcharts, APIs, and MCP. The buyer decision is whether business users can safely create useful automation without waiting for engineering—while IT retains permissions, approvals, audit logs, and deployment control.

## What it does

Jinba Flow turns a plain-language request into an inspectable workflow that can be edited visually, tested with real data, and published as an API or batch process. Jinba App lets colleagues execute the workflow through chat or generated forms; Jinba Toolbox provides a versioned tool registry and sandbox execution. The platform lists on-prem deployment, SSO, RBAC, audit logs, 100+ integrations, and private model options ([Jinba homepage](https://jinba.io/); [Jinba docs](https://docs.jinba.io/en/introduction)).

| Fact | What the public sources say |
| --- | --- |
| Buyer | Enterprise business teams, IT, operations, finance, and regulated organizations |
| Workflow | Describe, build, test, deploy, execute, monitor, and govern AI workflows |
| Public pricing | Free plan with 2 team members, 2 workspaces, 10 workflows, and 1,000 API credits; Enterprise custom |
| Controls | On-prem/private cloud, SSO, RBAC, audit logging, approvals, and private models are listed |
| Founders | Shoya Matsumori and Takuya Norisugi |

## Why it fits

The product is designed around the gap between “business users know what should happen” and “engineering must build every automation.” Chat is the entry point, but the important control is the visible flow, test run, approval policy, and reusable deployment. That makes it more credible for enterprise work than a prompt that silently calls tools.

The tradeoff is governance at scale. A buyer should resolve who can create a workflow, how secrets and connectors are isolated, how changes are versioned, how failed runs recover, and whether users can distinguish an AI step from a deterministic one. Jinba's homepage says SOC 2 compliant and offers on-prem deployment; those claims should be validated in procurement. Pricing beyond Free is custom.

The founder context fits: YC describes Matsumori as a CS PhD and Norisugi as a former McKinsey manager ([YC company profile](https://www.ycombinator.com/companies/jinba)). The launch says Jinba serves users in regulated industries; that remains company-reported.

Short version: Jinba is worth a pilot for one cross-app workflow with visible approvals and rollback. Let the business team describe the goal; keep production permissions with IT.

## Sources checked — 2026-09-19

- [YC company profile](https://www.ycombinator.com/companies/jinba)
- [Jinba homepage](https://jinba.io/)
- [Jinba pricing](https://jinba.io/pricing)
- [Jinba docs](https://docs.jinba.io/en/introduction)

## Cohort context

Jinba is listed in Winter 2026. In our 2026-09-18 directory snapshot, 126 of 199 listed companies in that cohort have YC’s primary industry label B2B (63.3%). This is a current-directory comparison, not an original intake count or a performance ranking. [Nine-cohort dataset](https://mudpie.ai/research/yc-cohorts-2026-09-19.json).

## Public website snapshot

Observed 2026-09-19T16:20:15.178Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.

| Signal | Homepage observation |
| --- | --- |
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Observed |
| Typed structured data | Observed |
| Docs/developer link | Observed |
| Pricing link | Observed |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |

[Public observations](https://mudpie.ai/research/yc-homepage-links-2026-09-19.json) · [Collection method](https://mudpie.ai/research/yc-homepage-methods/README.md). Missing links here do not establish that a capability or file is absent elsewhere.


## Author disclosure

I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.