AIエージェントと社内システムを、企業内MCPで安全につなぐ
既存API・Microsoft 365・社内データを、Copilot / OpenAI / Claude / 自社AIから安全に呼び出す企業内MCP基盤を設計・開発。接続方式を標準化し、認証・認可・監査を効かせたまま生成AIを実業務へ組み込みます。
- MCPサーバー構築
- 認証・認可・監査
- 既存API連携
- Microsoft 365
SmartDesigner
Which AI for which job? We design your AI lineup and budget.
We build an AI agent foundation with enterprise MCP, then assign high-performance, cost-efficient, and evaluator AIs by role - designing how AIs work together before adding headcount.
We clarify which AI handles what, and how much it costs, including AI budget estimates. Beyond PoC, we design authentication, permissions, and operations.
Agent foundationConnect internal systems and AI securely via enterprise MCP
Optimal AI placementAssign high-performance, cost-efficient, and evaluator AIs by role
AI budget designWhich AI at what cost - estimate and optimize ROI
Areas we can help with
- 企業内MCPサーバー構築
- AI Project Management（AI配置・予算設計）
- AIエージェント業務組み込み
- Microsoft Fabric（データ分析基盤）
- Data Agent（社内データ活用）
- Azureインフラ設計・構築
- ゼロトラストセキュリティ
- Azure Virtual Desktop
- AIロールプレイ研修
Hot Topics
どの仕事を、どのAIに。AIモデルの最適配置を設計します
高性能AI・コスト効率AI・評価AIを機能ごとに使い分け、役割分担・品質評価・コスト管理まで設計。「AIを使う」から「AI同士を役割分担させ、AIの判断をAIが評価する」体制づくりを支援します。
- AIモデル選定
- 役割分担設計
- AI品質評価
- コスト最適化
AI顧客と話して、営業・接客トークを鍛える
AIを相手にロールプレイし、会話後にAIが採点。改善点と模範回答を提示します。属人化しがちな営業・接客スキルを、繰り返し練習で底上げします。
- 営業研修
- 接客トレーニング
- AI採点
- 模範回答
社内データに、話しかけて答えを引き出す
自然言語の質問から、社内データを根拠付きで分析・可視化するData Agent。担当者に頼らず、必要な数字とインサイトにその場で辿り着けます。
- 自然言語分析
- 根拠付き回答
- 可視化
- 社内データ活用
どこからでも安全に。クラウドデスクトップ導入支援
Azure Virtual Desktopで、社内PC環境をクラウドへ。セキュリティを保ちながら、リモートワークや拠点分散に対応した仮想デスクトップ基盤の設計・構築を支援します。
- リモートワーク
- セキュリティ
- Azure
- 運用設計
社内データを一つの基盤へ。分析とAI活用の土台をつくる
散らばった業務データをMicrosoft Fabricに集約し、レイクハウスからPower BI・AIまでを一気通貫で整備。企業内MCPやData Agentが根拠として使える、信頼できるデータ分析基盤を設計します。
- データ統合
- レイクハウス
- Power BI
- 分析基盤
拡張に強い土台を。Azureインフラ設計・構築
ランディングゾーンの考え方に沿って、ネットワーク・ID・ガバナンスを標準化。小さく始めて安全に拡張できる、AIワークロードにも耐えるAzureインフラの設計・構築を支援します。
- ランディングゾーン
- ネットワーク設計
- ガバナンス
- 拡張性
「社内だから安全」をやめる。ゼロトラスト設計
Microsoft Entraを軸に、ID・デバイス・アクセスを常に検証するゼロトラストを設計。生成AIや外部連携が増えても、最小権限と可視化で守れる構成を整えます。
- Microsoft Entra
- 最小権限
- 多要素認証
- アクセス可視化
What We Do
We choose practical AI use cases together, tailored to your goals.
What matters in AI adoption is not starting from technology names, but identifying where operations take time and which data can be used to confirm results. SmartDesigner supports consultation, demos, PoC, and production in one continuous flow.
Reduce inquiry handling
Automate service explanations, FAQs, and initial interviews with AI avatars and AI chat, creating a path that leads to consultations.
See the AI avatar demoFind internal information quickly
Build RAG and knowledge search that answers with evidence, targeting PDFs, FAQs, SharePoint, and business databases.
Consult about internal searchStreamline data analysis and inspection
Use Microsoft Fabric, machine learning, and image recognition to support sales and inventory analysis, forecasting, and checking tasks.
Watch the Data Agent video
Demo Movie
Data Agent and AI Roleplay demo videos
Watch two demos: asking sales and inventory data questions in natural language with Microsoft Fabric Data Agent, and AI Roleplay Training where you practice sales and customer-service talks against an AI customer.
Check sales and inventory through conversation.
This is a demo of checking things like "How were last month's sales?" and "Is there any imbalance in inventory?" in natural language. You can consider adoption including data preparation, permissions, answer quality, and Teams integration.
Loading the video.
AIロールプレイ トレーニング
AI顧客を相手に営業・接客トークを練習。会話後にAIが採点し、改善点と模範回答を提示します。
Service Areas
Service areas
Based on your existing systems and internal data, we design configurations centered on Azure and Microsoft services. We start with a small PoC and gradually add the authentication, permissions, logging, and monitoring needed for production.
RAG and internal knowledge search
Build AI search that answers with evidence, targeting internal documents, FAQs, product information, and business databases.
- Azure OpenAI / Azure AI Search
- 引用表示、権限制御、監査ログ
- 回答品質の評価と改善
Microsoft Fabric and data utilization
Organize sales, inventory, and customer data and consider an analytics foundation usable with Power BI and Data Agent.
- Lakehouse / Warehouse / Data Agent
- Teams連携、権限、更新フロー
- 部門展開とガバナンス
AI agents and business integration
Connect existing systems, email, Teams, and business APIs to support review, drafting, and notification.
- Azure AI Foundry Agent Service
- 人間承認、最小権限、操作ログ
- 読み取り中心の安全なPoC
Machine learning and prediction models
Incorporate demand forecasting, classification, anomaly detection, scoring, material property prediction, and more into your operations.
- 目的変数、特徴量、評価指標の整理
- Azure Machine Learning / MLflow
- 再学習、監視、説明可能性
Image recognition and brand monitoring
For images and web pages, reduce the burden of visual inspection, detect suspicious expressions, and streamline checking tasks.
- 画像解析、生成AI、ルール判定
- ログ、判定理由、再確認フロー
- 運用しやすい管理画面
AVD and secure work environments
Design cloud work environments that let contractors, external members, remote workers, and GPU users work safely.
- Azure Virtual Desktop / Entra ID
- 条件付きアクセス、FSLogix、監視
- コストと運用手順の整理
Our SaaS Products
Our in-house SaaS products
Beyond contract and hands-on support, we also develop and provide our own SaaS that you can use right away in the field. You can start with core features for free, and we also support Azure deployment and operations.
営業・接客研修 / AIロールプレイ
For restaurants / QR ordering
SmartDesigner QR Order System
Smartphone ordering and checkout requests via QR codes, plus a new avatar ordering experience with animated avatars and voice input. Order, billing, and sales analytics in a single admin screen.
View product page
For SMBs / ERP and core business
SmartDesigner ERP
An ERP for SMBs that connects sales and billing, inventory, purchasing, approvals, and accounting in one. Visualize your management status with dashboards.
View product page
Industry-specific CRM / customer and deal management
SmartCRM
Centrally manage customers, companies, contacts, deals, activity history, inquiries, and tasks. With industry-specific field definitions, you can build a CRM tailored to industries such as manufacturing, IT, healthcare, and retail.
View product page
Simultaneous translation / inbound support
Smart Translation
Simultaneous translation that works with a single smartphone. It supports inbound customer service at stores and conversations on overseas business trips.
View product page
Joint project with a tax firm / OCR and accounting AI
ensidex AI Accounting Platform
A business support SaaS that integrates receipt OCR, expense settlement, invoice creation, and an accounting AI agent.
View product page
AI customer service / web concierge
AIコンシェルジュ りつ
An AI concierge you can embed in your website. It provides FAQs, facility guidance, and product guidance with a character UI.
See Ritsu's page
Education & Training / Workshops
Workshop Delivery for the AI Era
Done-for-you corporate training across a wide range—consulting, logical thinking, operational efficiency hackathons, AI, and cloud. We deliver hands-on workshops customized to your team.
See training details
How We Start
First, we break it into a small, testable scope.
Rather than lining up exaggerated numbers as achievements, we clarify what can be confirmed in public demos, what can be organized in the first consultation, and what should be validated in a PoC. We compile the materials needed for decisions into a form usable for sales documents and internal explanations.
01
Confirm the challenge
Organize the target operations, users, current burden, and expected effects.
02
Confirm usable data
Confirm the availability of PDFs, FAQs, SharePoint, SQL, images, logs, and more.
03
Decide the PoC scope
Narrow down to screens, data, and metrics that can be judged in a short time.
04
Consider production
Design including authentication, permissions, logging, monitoring, operations, and cost.
Pricing Guide
Pricing guide
To ease your concerns before consulting, we list rough guides for common support scopes. Actual costs are adjusted based on target data, existing environment, security requirements, and deadlines.
Free 30-minute consultation
FreeOrganize the challenge, target data, feasibility, and next step.
Technical challenge assessment
USD 600–1,800Organize the configuration proposal, PoC scope, estimate, risks, and preparation items.
RAG / AI chat PoC
USD 3,600–9,500Build data ingestion, search, answer UI, logging, and evaluation viewpoints.
AI avatar adoption PoC
USD 6,000–14,000Handle character, conversation design, knowledge integration, and web embedding.
Azure AI production support
From USD 12,000Cover authentication, permissions, monitoring, logging, security, and operational improvement.
Even if you're not sure which menu fits, we can organize it in a 30-minute consultation.
Consult about services and pricing
FAQ
Frequently asked questions
You can consult us even before requirements are finalized. First, we confirm feasibility, the data needed, and how to proceed with a PoC.
Can I consult even if I don't know where to start?
Yes. We review your business challenges, available data, and existing systems, and organize the first theme to try.
Can we start small?
Yes. By narrowing down to one theme, one dataset, and one business screen, you can confirm effects and risks in a short time.
Can it integrate with existing systems and Microsoft 365?
We design on the premise of integration with APIs, SQL, SharePoint, Teams, and Azure services, including permissions and logging.
Contact
Shall we organize your AI adoption approach in 30 minutes?
After seeing the demos, you can confirm which operations your company can start with AI, what data you can use, and the scope and cost needed for a PoC.
- You can consult even before requirements are finalized.
- You can consult concretely after seeing the demos.
- You can organize the PoC scope, cost, and required data.
- You can also consult about presentation that leads to inquiries.