AI Accounting & Finance Agent Platforms covers the growing ecosystem of agentic LLM tools, developer frameworks, automation platforms, and governance layers used to automate bookkeeping, financial analysis, reporting, and regulatory workflows. By 2026 these systems combine retrieval-augmented generation (RAG), stateful agent orchestration, model hosting, and curated training data to turn invoices, contracts, statements and PDFs into actionable accounting workflows while preserving auditability and compliance. Key building blocks include agent frameworks and marketplaces (for building, testing and deploying agents), document and RAG tools that surface facts from unstructured records, data‑curation services to prepare training data, and enterprise governance to validate models and vendors. Representative technologies: LangChain for engineering, debugging and deploying reliable agentic LLM applications; LlamaIndex for turning unstructured content into document agents and production RAG pipelines; PDF.ai for conversational access to PDF-based financial documents; Vertex AI and Google Gemini for managed model training, fine-tuning and inference at scale; DatologyAI for automated data curation and model-ready datasets; Mistral AI for efficient open models optimized for on‑prem or hybrid deployments; and Monitaur for governance, monitoring and policy controls in regulated contexts. This topic is timely because finance organizations are moving from manual and siloed automation to integrated agent platforms that must satisfy audit trails, privacy, and vendor governance while leveraging large multimodal models and curated data. Evaluations now focus less on novelty and more on reliability, explainability, data lineage, and operational controls — criteria that determine whether an accounting or finance AI agent can be adopted safely at scale.