Key facts
| API compatibility | OpenAI-compatible chat, embeddings and function calling |
| Use cases | Outage comms drafting, field procedure Q&A, filing support, billing enquiry handling |
| OT boundary | Keep control-system and SCADA data out of prompts; integrate by reference only |
| Data controls | Scoped API keys, RBAC, SSO, audit logs and region selection |
| Deployment | Plugsky cloud, your VPC, on-prem and air-gapped options |
| Models | 30+ models from free aliases to frontier reasoning |
| Pricing model | Flat monthly self-serve plans with unlimited fair-use usage |
| Free tier | plugsky-micro and plugsky-lite on the free plan, no card required |
TL;DR
- Draft outage and customer communications in a consistent, reviewable voice.
- Give field crews instant answers from procedures with citations.
- Speed up filing preparation while keeping legal review in the loop.
- Keep OT and SCADA data out of prompts entirely.
- Deploy on-prem or air-gapped where critical infrastructure policy requires it.
How it works, step by step
- Pick one workflow with a clear review owner, such as outage messaging or procedure Q&A.
- Set the data boundary first: no OT, SCADA or control-system data in prompts.
- Index approved procedures, manuals and public communications with stable document ids.
- Add citation requirements and a review step before anything is published or filed.
- Configure scoped keys, audit logging and the processing region to match policy.
- Pilot with one region or business unit, then expand after measuring accuracy and review time.
Try it yourself
Open the data residency checker →
Utility use cases worth starting with
Utility AI wins come from writing and retrieval, not control:
- Outage communications: draft consistent customer and media updates from incident facts for a duty officer to approve.
- Field knowledge: answer procedure, safety and asset questions from approved manuals with citations.
- Filing support: assemble first drafts of regulatory sections from prior filings and data extracts.
- Customer enquiries: classify billing and service questions and draft responses for agents.
Every output leaves through a named human owner.
Keeping OT out of the AI path
The hard line in utilities is between IT and OT. The AI layer should sit on the IT side and never receive control-system, SCADA or protection data in a prompt. Integrate by reference, sending incident summaries or document text rather than live telemetry. Within that boundary, configure scoped API keys per system, RBAC and SSO for staff, and audit logs so every request has an owner and a timestamp. Choose the processing region deliberately, and use a VPC, on-prem or air-gapped deployment where critical infrastructure policy or a regulator's expectations require stronger isolation.
Reliability, review and records
Utilities are judged on reliability and on the quality of the record. Keep the model out of automated control loops, require citations for any factual claim, and store the model version, prompt and retrieved sources alongside published output. That record supports after-action reviews and regulatory questions. Route classification and routine drafting to small models and reserve frontier models for complex synthesis such as multi-document filings. Because Plugsky is OpenAI-compatible, the integration tested in the cloud runs unchanged on-prem or air-gapped, which avoids parallel development paths.
A measured rollout
Prototype on the free plan with plugsky-micro or plugsky-lite on a single workflow and an approved document set. Use the 14-day full-access trial to compare models on real, sanitised cases, scoring factual support, review minutes and escalation rate. Self-serve plans are flat monthly with fair-use usage, so storm-season volume does not create billing swings. Roll out by business unit, keep prompts versioned, and re-run the evaluation set whenever procedures or models change. Expand only where the review process is already stable.
Honest comparison
| Capability | Plugsky | Typical per-token API | Building in-house |
|---|---|---|---|
| API compatibility | OpenAI-compatible chat, embeddings and tools | Usually compatible | Full rewrite |
| OT separation | AI stays on the IT side; integrate by reference | Depends on your design | You design the boundary |
| Deployment | Cloud, VPC, on-prem and air-gapped | Mostly cloud-only | You operate GPUs and serving |
| Pricing | Flat monthly self-serve, fair-use usage | Per-token, harder to forecast | GPU plus operations cost |
| Model choice | 30+ models behind one API | Varies by provider | You host every model |
Frequently asked questions
Can we keep our existing OpenAI SDK code?
Yes. Plugsky is OpenAI-compatible, so the change is a base URL and model name in code your team already maintains.
Is there a free plan?
Yes. The free plan includes plugsky-micro and plugsky-lite with no credit card, enough to prototype procedure Q&A or message drafting.
Does the AI touch SCADA or control systems?
No. Keep control-system and telemetry data out of prompts and integrate by reference; the AI layer stays on the IT side.
Can it write regulatory filings directly?
It can assemble drafts from approved sources, but legal and regulatory review must remain with accountable staff before anything is filed.
Can we deploy air-gapped?
Yes. Air-gapped deployment is available for critical infrastructure environments where no traffic may leave the site.
How does pricing work?
Self-serve plans are flat monthly with unlimited fair-use usage. See the live pricing page for current plans and enterprise options.
Which endpoints are live today?
Chat, streaming, JSON mode, function calling, embeddings, RAG and agents are live. Audio, images, moderation, files, batch, fine-tuning, assistants and responses are coming soon.