Forge turns private AI infrastructure into operator-ready workflows.
Deploy an operator-owned AI node on infrastructure you control, over data you own. Forge packages retrieval, inference, tool boundaries, evidence export, and observability into a last-mile system for real operations — without sending prompts or context to third-party AI services.
We start with renewable energy operations, proven in the field at SingingWater Vineyards. From there, Forge expands into other evidence-heavy workflows only where an operator-owned deployment model actually matters.
Deployment Layer
Forge Private AI Node — the last-mile private AI layer
Deploy a bounded, operator-owned AI node on-premise, in customer cloud, or at the edge. Connect documents, tickets, and approved workflows so your team gets consistent answers with citations — without standing up a generic AI platform program.
- • Retrieval, inference, and evidence export under your control.
- • Citations, guardrails, and access control over internal evidence.
- • Stable APIs and thin UIs for controlled adoption.
Renewable Operations
Forge Energy — the first vertical running on Forge.
Use Forge on renewable assets to quantify performance gaps and avoided loss, prioritize corrective action, and plan lifecycle decisions. Start with an operator-grade performance audit, then extend into OpenRoper field workflows and portfolio roadmaps.
- • Performance Audits grounded in real site data.
- • Forge Co-Pilot for technicians and operators for fault clarity, faster diagnosis, work orders, and truck-roll decisions.
- • Strategic use-case mapping for your portfolio and operating model.
Operator-owned AI nodes for real operations — deployed in customer cloud, on-premise, or at the edge where the work happens.
What is Forge?
A productized, operator-owned deployment layer — sold outcome-first and deployed behind customer-controlled boundaries.
Forge is not “AI consulting,” and it is not a blank AI platform. It is a standardized deployment that packages private inference, retrieval, tool boundaries, evidence export, and observability into something operators can actually use. Start with Forge Private AI Node as the foundation, then move into Forge Energy as the first verticalized outcome path.
Track 1 · Foundation layer
Forge Private AI Node
Deploy a bounded private AI node over internal evidence, SOPs, runbooks, tickets, and approved workflows — with predictable outcomes and no platform sprawl.
- • Retrieval, inference, and evidence export on your infra.
- • Citations and guardrails over internal evidence.
- • Stable APIs + thin UI so it ships as a product, not a science project.
Track 2 · Renewable operations first
Forge Energy
Use the same Forge foundation for renewable operations: performance gaps, avoided loss, and lifecycle decisions — starting with an operator-grade performance audit.
- • Operator Performance Audit grounded in site data.
- • Repeatable reporting and decision narratives.
- • Expand into copilot workflows.
Expansion
Forge Energy Co-Pilot
Forge is the foundation. OpenRoper is how operators and field teams use it: fault clarity, work orders, and technician-ready answers tied to real SOPs and site context.
- • Compass → Blueprint → Trailmap → FieldAssistant → Strategy Builder.
- • Field-first workflows for MTTR reduction and avoided loss.
- • Built to be operator-owned and audit-friendly.
Mid-market model: land with Forge Private AI Node ($10k/year or $850/mo), go live via a Launch Pack, then expand with support + modules.
What you get in 30 days
This is not “AI consulting.” It’s a repeatable deployment that ships copilots into real workflows, with audit-ready telemetry and strict data boundaries.
Clear promise
Forge goes live in 30 days with 3 operational copilots running on your infrastructure, over your data — no third-party prompt sharing required.
Guarantee
30-Day Go-Live Guarantee: Forge running + 1 copilot live, or you don’t pay the Launch Pack.
Value stack
- • Forge Node deployed (vLLM + Redis + MinIO + MQTT + Prometheus/Grafana)
- • 3 copilots shipped from templates (FieldAssistant / Fault Clarity / Strategy Builder)
- • Governance pack (data boundaries, retention, logging, audit trail)
- • Compass → Blueprint → Trailmap templates (your methodology, standardized)
- • Evaluation harness (baseline + drift monitoring path)
What we won’t do (on purpose)
- • No bespoke agent sprawl without a template spec + eval tests + priced module
- • No “send your data to a third-party LLM” architecture by default
- • No vague deliverables — everything ships as an operational workflow
Pricing built for mid-market
Land at $10k/year. Expand with Launch + Support + Modules.
Forge Core
Subscription (land)
$10,000 / year
- • Forge Stack license (operator-owned deployment)
- • Base model pack + upgrade cadence
- • Observability baseline (metrics + dashboards)
- • Quarterly product updates
Forge Launch Pack
One-time (standardized implementation)
$25,000 one-time
- • Go-live in 30 days: Forge + 1 copilot live (guarantee)
- • 3 copilot templates shipped (FieldAssistant / Fault Clarity / Strategy Builder)
- • 1 data connector (CMMS OR shared drive OR SCADA export)
- • Governance pack + success metrics baseline
Forge Ops Support
Expansion ARR
$12k–$30k / year
- • Patch cadence + security hardening support
- • Evaluation harness + drift monitoring path
- • Incident response SLA options
- • Roadmap priority for top customers
Optional modules
Connector Packs
+$6k/yr each
SharePoint, ServiceNow, Maximo, Zendesk, SCADA exports, more.
FieldAssistant Edge Mode
+$10k/yr
Offline-first patterns for jobsite use cases.
Prevention Credits Analytics
+$15k/yr
Avoided loss + uptime preservation narratives and reporting.
Delivery timeline: 4 weeks
Standardized delivery prevents custom-work sprawl and keeps outcomes predictable.
Week 1
Compass + readiness
- • Workflows + success metrics
- • Data boundaries + retention decisions
- • Environment readiness checklist
Week 2
Forge Node deployed
- • vLLM + model pack
- • Redis + MinIO + MQTT ingest
- • Prometheus/Grafana dashboards
Week 3
Copilots #1 and #2 shipped
- • Template install + tuning
- • Evaluation baseline captured
- • Stakeholder review + quick wins
Week 4
Copilot #3 + handoff
- • Governance pack
- • Operational runbook
- • Go-live support and rollout plan
FAQ
Do prompts or context get sent to third-party AI services?
Not by default. Forge is designed for operator-owned deployment. You control where inference runs and what data is accessible.
Is this “AI consulting”?
No. Forge is a productized deployment with repeatable templates. Anything outside templates becomes a paid discovery that produces a new template spec + eval tests.
What counts as “go-live” for the guarantee?
Forge Node deployed + observability live + 1 copilot live over your data source (or a provided sample dataset if internal access isn’t ready).
What does mid-market mean here?
Fast sales cycles, standardized deployment, and real operational impact without enterprise procurement drag.