--- title: "Proposal, tender and applied AI glossary" description: "Plain-language definitions for RFPs, tenders, proposal operations, AI agents and process automation." canonical: "https://zephior.com/glossary" last-updated: 2026-08-21 --- # Proposal, tender and applied AI glossary Plain-language definitions for RFPs, tenders, proposal operations, AI agents and process automation. Terms in this field are often used loosely. Each definition explains the operational meaning, the decision it affects and the nearby terms people commonly confuse. ## Published resources - [AI agent: meaning, architecture and production controls](https://zephior.com/glossary/ai-agent): An AI agent pursues a goal through model-guided decisions and tools, requiring explicit limits on identity, actions, state and human approval. - [AI guardrails: controls, limits and production testing](https://zephior.com/glossary/ai-guardrail): AI guardrails constrain inputs, outputs and actions, but no single filter guarantees safety. Design layered controls around the actual product consequence. - [AI model evaluation: test sets, metrics and gates](https://zephior.com/glossary/ai-model-evaluation): AI evaluation measures whether a model-enabled system meets task, risk, latency and cost requirements on representative cases before and after launch. - [Answer library: governed proposal knowledge for reuse](https://zephior.com/glossary/answer-library): Learn how an RFP answer library combines approved wording, source evidence, applicability, permissions, ownership and review dates. - [Answer provenance: traceable sources for proposal content](https://zephior.com/glossary/answer-provenance): Answer provenance records where a response came from, how it changed and who approved it. Learn how to make proposal claims reviewable. - [Award criteria: how public tenders evaluate offers](https://zephior.com/glossary/award-criteria): Award criteria define how admissible public tender offers are compared. Learn how to trace weights, evidence and scoring into a controlled response. - [Bid or no-bid: a practical decision framework](https://zephior.com/glossary/bid-no-bid): A bid or no-bid process decides whether an RFP or tender deserves pursuit capacity after gates, fit, evidence, economics and risk are tested. - [Business process automation: scope, design and control](https://zephior.com/glossary/business-process-automation): Business process automation coordinates work across steps, people and systems. Learn how to define scope, exceptions, controls and measurable outcomes. - [Compliance matrix: tender requirements control guide](https://zephior.com/glossary/compliance-matrix): A compliance matrix maps every tender requirement to its source, owner, evidence, response location, status and final verification. - [Conditions of participation in public procurement](https://zephior.com/glossary/conditions-of-participation): Conditions of participation test supplier capacity for a public contract. Learn to verify each condition, document and relied-on entity before bidding. - [Contract notice: finding and qualifying public tenders](https://zephior.com/glossary/contract-notice): A contract notice announces a procurement, but it is not the complete tender pack. Learn which notice fields matter and what suppliers must verify. - [CPV code: how European tender classification works](https://zephior.com/glossary/cpv-code): CPV codes classify the subject of European public contracts so authorities can publish and suppliers can search tender notices consistently. - [Due diligence questionnaire: DDQ meaning and process](https://zephior.com/glossary/ddq): A due diligence questionnaire helps a buyer or investor assess an organisation’s controls, risks, capabilities and evidence before a decision. - [E-procurement: electronic tendering from notice to contract](https://zephior.com/glossary/e-procurement): E-procurement moves notices, documents, questions, submissions and transactions online. Learn the supplier controls that prevent portal failure. - [Embeddings in AI: meaning, retrieval and evaluation](https://zephior.com/glossary/embedding): Embeddings turn content into numerical representations for comparison and retrieval. Learn what determines useful similarity in a production system. - [Exception handling in automated business processes](https://zephior.com/glossary/exception-handling): Exception handling detects cases that cannot follow the normal path and routes them to recovery. Learn how to design ownership, evidence and resolution. - [Fine-tuning AI models: meaning, choices and controls](https://zephior.com/glossary/fine-tuning): Fine-tuning changes a pretrained model for a defined task or behavior. Learn when it helps, what evidence it needs and how to operate it safely. - [Framework agreement: call-offs, value and bidder risk](https://zephior.com/glossary/framework-agreement): A framework agreement sets terms for future contracts, but supplier appointment may not guarantee orders, volume or the advertised ceiling value. - [Human in the loop: meaningful oversight for AI](https://zephior.com/glossary/human-in-the-loop): Human-in-the-loop AI places a person at a defined decision point, with enough evidence, authority and time to review or change the outcome. - [Intelligent document processing: IDP system guide](https://zephior.com/glossary/intelligent-document-processing): Intelligent document processing turns variable files into validated structured data through capture, classification, extraction and review. - [Large language model: LLM meaning and production use](https://zephior.com/glossary/large-language-model): A large language model predicts and generates language, but a reliable AI product also needs context, controls, evaluation and software operations. - [Model Context Protocol: tools, resources and boundaries](https://zephior.com/glossary/model-context-protocol): Understand MCP clients, servers, tools, resources and prompts, plus the security and reliability controls an enterprise integration still needs. - [Model drift: detection, diagnosis and response in AI](https://zephior.com/glossary/model-drift): Model drift is a production change that can invalidate expected AI behavior. Learn how to monitor outcomes, find causes and respond with control. - [Process mining: event data, discovery and action](https://zephior.com/glossary/process-mining): Process mining reconstructs how work actually flows from timestamped system events, revealing variants, delays, rework and control gaps. - [Prompt injection: risks, examples and defenses](https://zephior.com/glossary/prompt-injection): Prompt injection occurs when untrusted instructions alter intended model behavior, especially when an LLM can access sensitive context or tools. - [Proposal management: process, roles and controls](https://zephior.com/glossary/proposal-management): Proposal management coordinates qualification, requirements, writers, evidence, pricing, reviews and release into one controlled response. - [Proposal response: from buyer question to approved answer](https://zephior.com/glossary/proposal-response): A proposal response answers a buyer’s stated need with a controlled claim, evidence and commitment. Learn how teams build and review it. - [Public procurement: process, principles and bidders](https://zephior.com/glossary/public-procurement): Public procurement is the regulated purchase of works, supplies and services by public bodies through defined procedures, evidence and award rules. - [Response compliance: controlling RFP and tender answers](https://zephior.com/glossary/response-compliance): Response compliance means satisfying the buyer’s applicable instructions and requirements. Learn how to trace obligations and release with evidence. - [Retrieval-augmented generation: RAG system guide](https://zephior.com/glossary/retrieval-augmented-generation): Retrieval-augmented generation gives a language model selected external context, helping applications answer from current, private or cited sources. - [RFI: meaning, process and response strategy](https://zephior.com/glossary/rfi): An RFI, or request for information, helps a buyer understand supplier capabilities, solution options and market constraints before a formal procurement. - [RFP: meaning, process and practical use](https://zephior.com/glossary/rfp): An RFP, or request for proposal, asks suppliers to present a compliant solution, delivery approach, evidence and commercial offer for buyer evaluation. - [RFQ: meaning, quotation process and response guide](https://zephior.com/glossary/rfq): An RFQ, or request for quotation, asks suppliers to price a sufficiently defined requirement under stated commercial and delivery conditions. - [Robotic process automation: RPA meaning and fit](https://zephior.com/glossary/robotic-process-automation): Robotic process automation uses software bots to reproduce defined user actions, often bridging systems where stable APIs are unavailable. - [Standstill period after a public procurement award](https://zephior.com/glossary/standstill-period): A standstill period preserves time between an award decision and contract conclusion. Learn what suppliers should verify and do immediately. - [Straight-through processing: no-touch flow with control](https://zephior.com/glossary/straight-through-processing): Straight-through processing completes eligible cases without manual intervention. Learn how to define valid outcomes, exclusions and exception controls. - [Tender: meaning, procurement process and bidder guide](https://zephior.com/glossary/tender): A tender is a structured competitive procurement in which suppliers submit offers against defined requirements, rules and evaluation criteria. - [Tender clarification: how to ask and control questions](https://zephior.com/glossary/tender-clarification): Tender clarification is the formal route for bidders to resolve material ambiguity and incorporate authoritative answers before submission. - [Tender lot: scope, bidding rules and supplier decisions](https://zephior.com/glossary/tender-lot): A tender lot is a separately defined part of a procurement. Learn how lot scope, eligibility, award limits and dependencies shape a compliant bid. - [AI tool calling: schemas, permissions and safe execution](https://zephior.com/glossary/tool-calling): Tool calling lets a model request structured operations. Learn how applications validate arguments, authorize effects and verify real outcomes. - [Vector database: embeddings, search and production choices](https://zephior.com/glossary/vector-database): A vector database stores and searches embeddings by similarity. Learn what it does, when it helps and how to test relevance, filters, latency and change. - [Workflow orchestration: state, retries and recovery](https://zephior.com/glossary/workflow-orchestration): Workflow orchestration coordinates multi-step work across systems, people and AI while preserving state, policy, observability and recovery.