Skip to content

Case Study: Regulated Content Review (HITL)

Anonymous composite based on financial-services marketing workflows. Not a specific customer deployment.

Problem

A regional bank's marketing team wanted to use LLMs to draft product emails and social posts. Compliance requires human sign-off before any customer-facing text is stored or sent. Existing tools either:

  • Lacked audit trails, or
  • Required building custom interrupt logic on LangGraph/CrewAI.

Solution with HiveFlow

Architecture

flowchart LR
    U[Marketing user] --> S[HiveFlow Studio]
    S --> A[Agent plan-only]
    A --> P[Plan: research → draft → compliance → publish]
    P --> H[HITL gate: compliance]
    H --> R[Reviewer in Approvals UI]
    R --> E[execute-plan]
    E --> BB[Audited blackboard]
    BB --> Rep[Replay / Analytics]

Configuration

HIVEFLOW_RUNTIME=agent
HIVEFLOW_PLAN_HITL=true   # plan review before execution
# Production: real LLM providers; CI: HIVEFLOW_AGENT_ECHO_LLM=true

Workflow

  1. Plan-only — Agent proposes research → draft → compliance_review → final_answer.
  2. Plan HITL — Compliance officer edits plan JSON in Approvals (e.g. add mandatory legal_check node).
  3. Execute-plan — Approved graph runs; compliance_review node has per-node HITL with draft payload.
  4. Audit — Every blackboard write and gate decision keyed by intent_id; Replay page for regulators.

Outcomes (internal pilot metrics)

Metric Before (manual chain) With HiveFlow
Median draft-to-approved time ~2 days ~4 hours
Audit export for exam Spreadsheet + email Blackboard audit + Replay JSON
Engineer maintenance Custom Flask + scripts Studio + Core APIs

Pilot n≈12 workflows; not a published benchmark.

Why not LangGraph alone?

LangGraph supports interrupts, but this team needed:

  • Studio Approvals for non-engineer reviewers
  • Encrypted blackboard keys for draft content
  • Dual guards on ingest (PII patterns) and output (forbidden claims)

HiveFlow provided these without a separate LangSmith-style hosted tier.

Reproduce locally

  1. HITL cookbook + examples/03_hitl_approval.py
  2. Studio Agent mode with HIVEFLOW_PLAN_HITL=true
  3. Walk through Approvals → execute → Replay