# TechJapan LLC > Ontology platform development. TechJapan LLC designs the knowledge structure underneath an AI system — terminology, relationships, constraints — and builds the retrieval that runs on it. Based in Hadano, Kanagawa, Japan. Last updated: 2026-08-29. Canonical site: https://www.tech-japan.jp Prices exclude Japanese consumption tax and match the corresponding product pages. This file is generated from the same source as the website and /facts.json, so the three cannot disagree. ## Company - Legal name: TechJapan合同会社 (TechJapan LLC) - Corporate Number: 1011003018911 - Founded: 2023 - Representative: Go Kyono (興野 剛), Managing Member - Address: 1-45 Ohata-cho, Hadano, Kanagawa 257-0015, Japan - Service area: Japan, remote. Southeast Asia through partner companies - Working languages: Japanese, English, Thai - Public registry: https://info.gbiz.go.jp/hojin/ichiran?hojinBango=1011003018911 - Contact: https://www.tech-japan.jp/contact/ ## Machine-readable - [facts.json](https://www.tech-japan.jp/facts.json): every fact on this page as JSON, generated from the same source. - [llms-full.txt](https://www.tech-japan.jp/llms-full.txt): every published fact in one fetch, including per-product tables. ## Pages - [Overview](https://www.tech-japan.jp/): what we build, and how the retrieval works. - [About](https://www.tech-japan.jp/about/): company record, timeline, and the representative's background. - [Services](https://www.tech-japan.jp/services/): what can be bought, and at what price. - [Work](https://www.tech-japan.jp/works/): products in production, the award-winning project, and contract scope. - [Writing](https://www.tech-japan.jp/blog/): design notes and measured comparisons. - [News](https://www.tech-japan.jp/news/): awards, credentials, press. - [Contact](https://www.tech-japan.jp/contact/) ## Products ### Hadano AI Cabinet — in development A local, evidence-first knowledge database for AI agents: it finds the passage that backs an answer, then expands around it. - **Interface**: Seven MCP tools over stdio. No HTTP API, no SDK, no interface of its own - **Stage 1 retrieval**: Keyword search over a character-trigram index, ranked by BM25. No model, no network, no sampling - **Stage 2 retrieval**: A one-hop walk over the relation graph, seeded by at most three chunks that stage 1 actually matched - **Vector search**: Designed, not connected in this version. Asking for it returns an error rather than an empty answer that looks real - **Measured quality**: nDCG@10 of 0.58 on SciFact - **Storage**: One file on the machine. No server, no API key, no GPU - **Network**: None. The local build makes no outbound calls - **Requirements**: Python 3.11 or newer - **License**: Apache-2.0 - **Availability**: In development. Not on PyPI yet, so pip install does not work today - **Implementation**: Seven of nine specified MCP tools are built and reachable. The other two are designed, not built - Page: https://www.tech-japan.jp/products/hadano-ai-cabinet/ - Site: https://www.h-c.ai/ ### Pasha Kintai — in production since February 2026 Photograph a paper timesheet with a phone. AI-OCR reads the handwriting and generates the attendance record. - **Price**: JPY 100 per person per month, tax excluded. One plan - **Setup fee**: None - **Billing**: Usage based. Only users with at least one clock-in or OCR import that month are charged - **Example**: Ten active users is JPY 1,000 per month, tax excluded - **Free trial**: One month or longer, every feature, no cap on users or uploads, no credit card - **Minimum term**: None. A month with no usage costs nothing; the account can be paused or cancelled at any time - **Capabilities**: AI-OCR for handwritten timecards, a correction workflow with named approvers, labor-law-compliant aggregation, PDF, Excel and CSV export, and notification when the statutes change - **Who it is for**: Japanese manufacturers with roughly 10 to 50 employees - Page: https://www.tech-japan.jp/products/pasha-kintai/ - Site: https://pasha-kintai.com ### MiiHawk — in production Collects boarding data by NFC or QR scan on ordinary phones, then builds a minimum-cost dispatch plan with a mathematical optimization engine. - **Reported effect**: Up to 40 percent reduction in employee shuttle bus cost - **What is optimized**: Vehicle combination and fleet size. This is not GPS route guidance - **Hardware**: None. NFC and QR scanning on ordinary smartphones - **Offline**: Runs offline as a PWA and syncs when the connection returns - **Countries**: Thailand, Myanmar, Vietnam, Indonesia, Malaysia - **Interface languages**: Japanese, English, Thai - **Limited Edition**: THB 20 per person per month: data collection, master management, dashboard and six report types, real-time monitoring, offline support, automatic departure timetable - **Enterprise Edition**: THB 30 per person per month: adds dispatch simulation by mathematical optimization, shift linkage and automatic planning, five kinds of cost-reduction proposal, Excel and PDF output - **Enterprise plus Route Optimizer**: THB 100 per person per month: adds route consolidation, before and after comparison, Google Maps integration, and an advisor - **Free trial**: Two weeks, up to two routes and 3,000 registered employees - Page: https://www.tech-japan.jp/products/miihawk/ - Site: https://miihawk.com/ ## Services ### Ontology and knowledge graph construction For: Teams whose LLM or RAG deployment is not answering accurately enough - Knowledge asset inventory - Ontology design: classes, relations, constraints, controlled terminology - Knowledge graph construction on Neo4j or an equivalent store - Hybrid retrieval and semantic search implementation - Quantitative retrieval accuracy evaluation against a fixed gold set - Handover of the operating and maintenance process - **Price**: From JPY 30,000 per hour, tax excluded. First consultation free - Page: https://www.tech-japan.jp/services/ontology/ ### AI integration, hands on - **Strategy**: Inventory of operations and data assets, opportunity identification, agent design, use case selection, ROI estimate, roadmap. Subsidy application support - **Integration**: Agent construction, generative AI integration, data pipelines, production deployment, API integration, interface work, security and compliance - **Operation**: Monitoring, model improvement, team training, cost optimization, and handover so the client can run it without us - **Price**: JPY 30,000 per hour, tax excluded. No setup fee, no minimum hours. First consultation free - Page: https://www.tech-japan.jp/services/ ### kintone and AI - AI-OCR so paper reaches kintone without retyping - Generative AI and LLM integration - Data analytics and prediction - Agent automation across apps - Internal knowledge search over kintone data - External system and API integration - Standard kintone design, app development, customization and enablement - **Price**: Quoted per scope. Hourly from JPY 30,000, tax excluded. First consultation free - Page: https://www.tech-japan.jp/services/kintone/ ### Offshore development Partner companies implement. Requirements, design, quality control and acceptance testing stay with TechJapan in Japan. - Not a fit: Small builds of a few screens. For those we propose direct domestic development or kintone instead. - **Price**: Quoted by scale and team structure. The first consultation is free and includes a rough team plan and cost estimate - Page: https://www.tech-japan.jp/services/offshore/ ## Awards and credentials - **Legal Data Utilization Award, Digital Agency Law x Digital Hackathon 2025** (2026-03). Fukko Compass, a disaster-recovery navigation system. Team Dango, including Go Kyono, won for a system that reconciles scattered disaster-relief programs against legal data and tells a survivor which programs apply and what to do next. 112 participants across 20 teams; ten teams reached the final. Source: https://www.digital.go.jp/policies/legal-practice/6acf8ad5-3668-4db4-b998-1c20eff10017 - Digital Promotion Officer appointed by the Digital Agency of Japan - Member of the Japanese Society for Artificial Intelligence - Legal Data Utilization Award, Digital Agency Law x Digital Hackathon 2025 - Kanagawa Governor License (1) No. 32931 ## Writing - **Ontology design in practice: the knowledge structure that decides retrieval accuracy** (2026-08-24): Why retrieval accuracy problems come from unstructured knowledge, and a five-step method from terminology control to a working knowledge graph. https://www.tech-japan.jp/blog/ontology-design/ - **Legal Data Utilization Award at the Digital Agency Law x Digital Hackathon** (2026-04-01): Fukko Compass: P-A-B tagging, the patented Dango interface, and My Number Portal integration. https://www.tech-japan.jp/blog/hackathon-award/ - **A comparative study of text chunking methods** (2026-03-10): Semantic, recursive and fixed-length chunking compared for retrieval quality and context efficiency. https://www.tech-japan.jp/blog/chunking-research/ - **The real reason AI fails in accounting** (2025-02-20): The blocker is not model accuracy. It is that the business context behind a journal entry was never written down. https://www.tech-japan.jp/blog/accounting-ai-failure/ - **AI outcomes are decided by data design** (2025-01-15): Structuring low-volatility data is what compounds into a long-term advantage. https://www.tech-japan.jp/blog/ai-data-design/ - **Why AI adoption stalls at the proof of concept** (2025-01-10): Root causes in organizational design and data integration, and what the production-bound version has to contain. https://www.tech-japan.jp/blog/ai-poc-failure/ - **Why SaaS is reaching its limits in the AI era** (2025-01-05): The case for integration over subscription when the operation is specific to one company. https://www.tech-japan.jp/blog/saas-ai-limits/ ## Frequently asked **What does TechJapan LLC actually do?** We build ontology platforms: the knowledge structure underneath an AI system, and the retrieval that runs on it. That work reaches the world three ways, as our own products, as hands-on integration for clients, and as contract development. **Why does the ontology matter more than the model?** When a retrieval deployment answers badly, the cause is usually the structure of the knowledge, not the model. Terminology is inconsistent, relationships between concepts are undefined, and each department keeps its own version of the truth. No model choice repairs that. **How much does AI integration cost?** JPY 30,000 per hour, tax excluded, with no setup fee and no minimum hours. The first consultation is free. Japanese IT and DX subsidies may apply, and we help with the application. **Is Hadano AI Cabinet available today?** Not yet. It is in development. Seven of nine specified MCP tools are built and reachable, but the package is not on PyPI and the repository is not public. We keep what is built and what is available in separate columns on purpose. **Can a small company work with you?** Yes. Small and mid-sized businesses are the primary focus. An engagement normally starts with an inventory of existing operations and data, then a roadmap chosen for the lowest cost that still produces an effect. **How do you keep a project from stalling at the proof of concept?** The work is scoped in three phases on the assumption that the system reaches production: strategy, integration, then operation. Monitoring, an improvement cycle, and handover to the client team are inside the scope, not after it. **Which countries do you cover for offshore development?** Vietnam, Thailand and Myanmar, through partner companies. Requirements, design, quality control and acceptance stay with us in Japan. **Where are you, and where do you work?** Hadano, Kanagawa, Japan. Engagements run remotely anywhere in Japan, and we work across Southeast Asia through partners. **Is there a free trial of Pasha Kintai?** Yes. One month or longer, every feature, no cap on users or uploads, and no credit card. **Can Pasha Kintai be used only in busy periods?** Yes. Billing follows actual usage, so a month with no clock-in or OCR import costs nothing. There is a pause function, no minimum term, and cancellation at any time.