Terms that matter in energy ops and AI.
Clear definitions for solar EPC and AI agent concepts. These pages are built for search intent: low competition, high specificity.
Solar EPC Terms
AHJ
Authority Having Jurisdiction: the local entity that reviews and approves solar permits.
PTO
Permission to Operate: the utility's formal approval to energize a solar system.
NTP
Notice to Proceed: the formal signal that kicks off construction or a major project phase.
Closeout Package
The final set of documents delivered after construction completes.
Lien Waiver
A document that releases a contractor's right to file a lien after payment.
As-Built Documents
Drawings that reflect what was actually constructed, including any field changes.
Permit Expiry
The deadline by which a permit must have work started or it becomes invalid.
Inspection Result
The pass, fail, or conditional status returned by an AHJ inspector.
Daily Report
A field log of crew attendance, work completed, and issues encountered.
Phase Advance
Moving a project from one stage to the next upon completion of prerequisites.
Subcontractor Follow-Up
The repeated communication required to get documents or status from subs.
Milestone Invoice
Billing triggered by the completion of a defined project stage.
Bootstrap Agent
A lightweight agent built quickly to prove a workflow before scaling.
AI Agent Terms
Agent Orchestration
Coordinating multiple agents so they work together without conflicting.
MCP
Model Context Protocol: a standard for connecting LLMs to tools and data sources.
Heartbeat
The scheduled interval at which an agent checks for new work.
Shadow Mode
Running an agent alongside human processes without taking real actions.
Tool Use
When an LLM invokes an external function to read or write data.
Sub-Agent
A smaller agent delegated to handle a specific task by a parent agent.
AgentMail
An email-native interface for AI agents to send and receive messages.
Inbox Agent
An agent that monitors an email inbox as its primary input source.
Agentic AI
AI systems that can take autonomous actions to achieve a goal.
LLM
Large Language Model: the reasoning engine behind modern AI agents.
Webhook
A mechanism for one system to push real-time event data to another.
Prompt Engineering
Crafting instructions to get reliable, structured output from an LLM.
Context Window
The amount of text an LLM can consider at one time.
Agent Memory
How an agent retains information across conversations or tasks.
RAG
Retrieval-Augmented Generation: fetching relevant documents before generating a response.
Function Calling
An LLM capability to output structured requests for external tools.
Observability
Logging and tracing agent actions so you can debug and improve them.