Worth a look if you run a regulated Japanese or Korean enterprise and need document agents on your own servers. Its public listing is slated for early November 2026 under ticker code 656A, so public investor filings and quarterly results should start to appear for outside readers to check.
Allganize is an enterprise AI company that started in Oakland, California, and is now headquartered in Tokyo with an office in Seoul. It sells Alli Coworker, a plain-language builder for AI apps and agents, and Alli Works, a ready-made document agent, both deployable in the cloud, on-premises or air-gapped.
Founded: 2017 · HQ: Tokyo, Japan · CEO: Changsu Lee · Funding: About $35M across four rounds (Seed $1.6M 2018, Series A1 $3.4M 2019, Series A2 $10M 2021, Series B $20M 2023), per the company
About Allganize
Allganize was founded in July 2017 in Oakland, California, by Changsu Lee, Kibin Shin and Yasuo Sato. Lee holds bachelor's and master's degrees in computer engineering from KAIST, worked at SK Telecom and GameOn, then started the mobile game analytics startup 5Rocks and sold it to Tapjoy. Sato spent 20 years in business development and ran 5Rocks' Japanese business. Shin worked at Neowiz, Naver and Line on server backends and machine-translation voice recognition. The company says it began with natural language search, an AI task bot named Alli and AI consulting. It opened a Seoul office in May 2018 and a Tokyo office in January 2019, and moved its headquarters to Tokyo in December 2021.
Today the product line centers on Alli. Alli Coworker lets staff build AI apps and agents in plain language, deploy them on the web or inside Slack or Teams, and run them on-premises or air-gapped, with ERP and groupware connections. Alli Works is a ready-made document agent for contract review, RFP analysis and guideline checks, and Allganize says every answer cites the source passage. Alli Documents 2.0 handles document understanding and extraction, Deep Research produces sourced reports across multiple sources, and a desktop Works Agent processes Excel, PDF and PowerPoint files locally. The company states that SaaS and on-premises deployments run from a single codebase and that customers can choose their own language model. It competes for the same buyers as Glean, Writer and Microsoft Copilot Studio, and the models behind such apps can come from vendors like Anthropic or OpenAI.
Funding is modest next to the frontier labs. Allganize's own history page lists a $1.6 million seed in August 2018, a $3.4 million Series A1 in April 2019, a $10 million Series A2 in February 2021 and a $20 million Series B in November 2023, which adds up to about $35 million. It began preparing for an IPO in January 2023. According to Matsui Securities, Allganize Holdings (code 656A) is scheduled to list on the Tokyo Stock Exchange Growth market on 4 November 2026, with an offering price of 1,260 yen and 19,073,388 shares outstanding at listing, of which 658,100 are newly offered and 2,954,900 are sold by existing holders.
Leadership combines founders and hired executives. Lee is founder and CEO, Sato is co-founder and Japan country manager, and Shin is co-founder. Moonsik Kang is CTO, with earlier roles at Lyft, Tapjoy, 5Rocks, NCSoft and Naver. Kensaku Kobayashi is CFO and a Japanese certified public accountant who previously worked at KPMG AZSA. Junghoon Lee is chief AI officer, a former NLP modeler at Interpark who led the RAG and AI teams at Allganize. Jay Tchakarov runs the US business. The pages read for this record do not state a total headcount.
On its homepage Allganize says it is in production at more than 240 enterprise customers and shows logos including Nomura Securities, Tokyo Metro, AEON, ENEOS, NTT docomo, JAL and ANA. It lists industry pages for financial services, insurance, manufacturing, construction, oil and gas, human resources and IT. Its pitch is on-premises and air-gapped deployment plus document answers that point back to the source page. HokAI has not tested the vendor's accuracy claims. Its home market overlaps with Korean enterprise AI names on HokAI such as Upstage, Naver and Kakao, and with Japan's Sakana AI. Vendors that also sell private or on-premises deployments include Cohere and Mistral AI, and LG AI Research is another Korean group building enterprise models. Buyers weighing a governed in-house setup can compare it with the general assistants Claude and ChatGPT.
The homepage lists SOC 2, HIPAA and ISO 27001 badges. No public trust center or audit report was found while researching this page, so the certification types, dates and scope are unconfirmed. Treat them as vendor statements and ask for the reports before a security review.
Mission
We improve the lives of the world's workers through the power of AI.
Products
- Alli Coworker (No-code AI app and agent builder): https://ai-coworker.alli.ai/portal/landing
- Alli Works (Document AI agent): https://www.allganize.ai/en/product/alli-works
- Alli Documents 2.0 (Document understanding and extraction): https://www.allganize.ai/en/product/alli-documents
- Alli Deep Research (Multi-source research reports): https://www.allganize.ai/en/product/alli-deep-research
- Works Agent (Desktop) (Local desktop document agent): https://www.allganize.ai/en/product/alli-works-agent
Compliance
SOC 2, ISO 27001, HIPAA
Links
Frequently Asked Questions
How much funding has Allganize raised?
Allganize lists four rounds on its history page, starting with a seed of $1.6 million in 2018 and ending with the Series B of late 2023, which brought in $20 million, about $35 million in total. Investor names were not read on those pages. Matsui Securities reports an offering price of 1,260 yen per share ahead of a scheduled 4 November 2026 Tokyo listing.
What products does Allganize make in 2026?
The suite is built around Alli: Alli Coworker for building AI apps and agents in plain language, Alli Works for contract, RFP and guideline work on your documents, Alli Documents 2.0 for extraction, Deep Research for sourced reports, and a desktop Works Agent. Allganize points buyers to a demo form and a pricing page. No list prices were read for this record, so none are stated here.
How does Allganize handle data privacy and compliance?
Badges for SOC 2, HIPAA and ISO 27001 appear on the homepage, but without a trust center or audit report the report types, dates and scope are unconfirmed. Allganize says its software runs on-premises and air-gapped, which keeps data inside the customer's own environment.
Who are Allganize's main competitors?
Glean is the closer rival on enterprise search, with a larger Western customer base, while Allganize leans on Japanese and Korean deployments and on-premises installs. Microsoft Copilot Studio wins on reach inside Microsoft 365 shops, and Allganize's case is air-gapped use and model choice. Writer targets similar enterprise agent buyers, and Allganize has fewer public integrations to point to.
How do you start using Allganize's products?
Request a demo through the contact page, or open the Alli Coworker app from the product menu and build a first app in plain language. Documentation for Alli Works and the API is at docs.allganize.ai. Enterprise on-premises installs go through Allganize's sales team.