Use Cases

FinClaw targets mid-to-large enterprises building AI agents and digital-employee platforms — especially scenarios with strict requirements on data security, compliance auditing, and personalized governance in private deployments. Typical use cases:

1. Enterprise Knowledge & Digital Employees

Turn expert skills, business experience, and internal knowledge into reusable skill modules, building role-agnostic, tireless "digital employees."

• Forms: customer-service agent assistant, ops patrol helper, R&D knowledge Q&A, sales script coach.
• Typical scenarios: employees call AI via natural dialog to generate reports, look up SOPs, review contracts, and other repetitive work.
• Value: convert tacit knowledge into digital assets and reduce business disruption from staff turnover.

2. Office Collaboration Integration

Through the application interaction layer, FinClaw integrates deeply with existing office tools so AI agents fit seamlessly into daily workflows.

• Entry points: DingTalk, Lark, WeCom and other IM tools.
• Business systems: OA, ERP, CRM, HRM and other core systems.
• Usage: summon agents in familiar chat windows for approvals, order creation, customer follow-up, and more.

3. Proactive / Triggered Scenarios (Pulse / Flow B)

Unlike traditional Q&A AI that only reacts, FinClaw's Pulse flow supports proactive triggered Agent execution.

• Trigger sources: scheduled jobs, business events, external webhooks, MQ messages.
• Examples: anomalous-order auto alerts with remediation suggestions, key-account birthday outreach, compliance daily reports, proactive sentiment monitoring digests.

⚠️ The Pulse / Flow B services (Trigger Gateway, Trigger Scheduler, Event Listener, Agent Router, etc.) are still under development and are not available in the current release.

4. Multi-Tenant AI Governance

The native multi-tenant system gives each department / subsidiary / business line an isolated virtual AI space.

• Independent config: each tenant configures its own data scope, model routing, skill permissions, and security policies.
• Scenarios: group with multiple business units sharing AI capabilities but isolated data; SaaS-style AI middle-platform for multiple clients.
• Control view: a "god's-eye" control center for global manageability, unified audit, and quota management.

5. Complex Workflow Orchestration

For multi-step, multi-role business processes, DAG workflow orchestration and long-thread execution channels support complex tasks.

• Examples: due-diligence report generation, multi-round negotiation preparation, cross-system data reconciliation.
• Core capabilities: high-concurrency scheduling engine, Agent Loop closed-loop execution, infinite-loop detection and circuit-breaking.
• Value: prevents runaway tasks from blowing up compute cost; guarantees controllable task completion.

6. Strong-Compliance / Financial-Grade Scenarios

Rely on the financial-grade security system to meet regulated-industry compliance.

• Target industries: banking, brokerage, insurance, central/state-owned enterprises, government.
• Key capabilities:
  • Four-layer physical sandboxing + 16-layer defense-in-depth
  • End-to-end auditable, tamper-proof audit logs
  • Intent sanitization, information-flow taint tracking, egress leak prevention
  • Data stays in the domain; supports fully offline deployment
• Typical deployment: bank private cloud, isolated intranet, info-tech-innovation hardware/software stacks.

7. Mobile / Web AI Entry Points

Via the ChatKit SDK and Web conversation entry, integrate AI capabilities into proprietary apps and portals quickly.

• Mobile: iOS / Android / HarmonyOS apps integrate the ChatKit SDK for instant conversational AI.
• Web: embeddable conversation entry, with info-tech-innovation OS and browser support.
• With FinClip: combine with the FinClip SuperApp platform to fold the AI dispatch assistant into the mini-program ecosystem.

---

Deployment forms covered by this documentation: all scenarios above can be privately deployed via FinClaw's single-machine Docker or Kubernetes cluster approach. See Quick Installation and the Deployment Guide for details.