Lobby
Conversation and context gathering via Teams or chat before handoff.
Platform
How OrqForge orchestrates enterprise work
OrqForge transforms unstructured inputs — Teams messages, emails, meeting transcripts, webhooks — into structured deliverables tracked, reviewed by humans, and published to your systems. A self-learning loop continuously improves accuracy over time.
The execution pipeline
Every run in OrqForge follows the same governed pipeline. Whether triggered from Teams, email, a meeting recording, or a scheduled job — work is ingested, classified, planned when needed, executed by role agents, reviewed, published, and learned from.
This is not one general assistant trying to do everything. It is a coordinated set of trained agents, each aligned to a role and stream of work, operating with stage history, review gates, and retained evidence.
At a glance
Five stages that cover the full lifecycle from trigger to improvement.
- 1
Ingest
Email, Teams, meetings, schedules, and custom listeners.
- 2
Triage & Plan
Classify work and gather missing context when needed.
- 3
Execute
Agent network delegates and produces deliverables in parallel.
- 4
Review
Human or AI approval before anything material is published.
- 5
Publish & Learn
Deliver results and improve from every completed run.
The full eleven-stage pipeline
Every stage is visible, inspectable, and retained for audit.
Ingest
Accept input from email, meetings, webhooks, or manual submission.
Triage
Classify work against your taxonomy and assign a confidence lane.
Planning
Structured Q&A when more context is needed before execution.
Hierarchy
Create structured work items with clear ownership and scope.
Execute
Role agents produce deliverables — documents, code, reports, actions.
Post-Execute
Refine and shape agent output into a coherent deliverable.
Review
Human or AI review with configurable approval gates.
Publish
Deliver results to SharePoint, Jira, email, or your systems.
Learn
Extract lessons from reviewer corrections and planning patterns.
Complete
Close the run with full stage history and retained evidence.
Product
Agent interfaces
Workspaces, hierarchy, triggers, and pipeline visibility — explore all screens on the demo tour.
app.orqforge.com/workspaces/demo-insurance/ai-trainer
AI Trainer 19 messages
I've reviewed your process docs. For this insurance example I'm proposing 6 roles across intake, evaluation, adjustment, and publishing. Enable orchestrator + FNOL + damage eval for the demo?
Yes — enable those three plus adjuster. Keep fraud detection disabled for now.
Done. Taxonomy updated: claim_intake, damage_assessment, adjustment_report. Review gates set to draft mode.
Attach documents or describe changes…
Send
Configure workspaces, roles, taxonomy, and review paths in a guided session — days, not years.
app.orqforge.com/workspaces/demo-insurance/org/organisation
Agent hierarchy
Insurance claims (demo) · 6 roles · 3 tiers
1
Process Orchestrator
Stream owner · Ops Lead (human)
Draft execution
-
Intake Agent
Teams · Email · Webhook
-
Classification Agent
Taxonomy routing
2
Delivery Orchestrator
Parallel specialist coordination
-
Damage Evaluation
Scoped tools · Evidence pack
-
Adjuster Agent
Structured output · Review path
-
FNOL Agent
First notice · Complete
3
Review Coordinator
Human accountability gate
-
Stream reviewers
2 assigned · Draft mode
Mirror your operating model: orchestrators, specialists, scoped tools, and human accountability gates.
app.orqforge.com/workspaces/demo-insurance/triggers/triggers
Multi-channel intake
Meet teams where they work — same governed pipeline for every trigger
T
Microsoft Teams
ActiveBot installed · 3 workspaces
@
Email listener
Activeclaims-inbox@demo.example · Rule-based routing
W
Webhook
ActiveAzure DevOps · Jira · Custom
⏱
Scheduled jobs
ConfiguredDaily heartbeat · Weekly reports
Enterprise pattern
Imperfect input from any channel → classified → orchestrated → reviewed → published to your systems.
Teams, email, webhooks, and schedules — every intake path feeds the same governed pipeline.
app.orqforge.com/workflows/ai-workflows
Motor claim intake — CLM-2026-0847
claim_intake · Demo Insurance POC · Active · 12m 34s
Lobby
Ingest
Triage
Plan
Hierarchy
Execute
Post-Exec
Review
Publish
Learn
Done
Damage assessment report
Execute · 8m 12s · 24.1k tokens
Claims orchestrator
Execute · Delegating · 12m 34s
Live pipeline stages, elapsed time, token usage, and status on every run — inspectable end to end.
Agent delegation
Chain of command, not a single prompt
When an agent encounters work outside its skillset, it reaches into its network and delegates. In our insurance demo, a process orchestrator coordinates damage evaluation and adjuster agents in parallel — each doing its part while the orchestrator monitors status and reasoning. The same delegation model applies to any workflow you build.
- Each role agent carries scoped tools, review paths, and workflow types
- Orchestrators delegate and monitor; specialists execute in parallel
- Full reasoning visibility on every active run
- Delegation lineage retained for audit and inspection
Example · Insurance demo
Process Orchestrator
Delegates work · Monitors team status
Damage Evaluation Agent
Specialist · Parallel execution
FNOL Agent
Specialist · Parallel execution
Adjuster Agent
Specialist · Parallel execution
Fraud Detection Agent
Specialist · Parallel execution
Workspaces and organisation structure
Workspaces mirror your internal organisation — customer boundaries, product lines, business units, or process domains. Within each workspace, role agents are bound to specific accountability: which tools they use, which review path applies, and which workflow types they handle.
The AI Trainer is OrqForge's agentic flow builder. It helps you design this structure for any process — not only insurance. Provide documents, a website, or process material — and OrqForge proposes workspaces, agent roles, taxonomy, and skills. For many organisations, a complete structure emerges in a short guided session.
Input sources
- Teams chat — agents installed as Teams bots; users converse and hand off work
- Email inboxes — shared or user mailboxes with rule-based routing
- Meeting recordings — transcripts processed automatically after meetings end
- Webhooks — Azure DevOps, Jira, and custom integrations
- Scheduled jobs — heartbeats, reports, and recurring checks
- Manual ingest — operators submit content through the platform
See the platform in action
Book a live walkthrough tailored to your processes, roles, and review requirements.