This topic examines the software and platform landscape that powers modern in‑vehicle AI—driver assistance, agentic features (autonomous or semi‑autonomous agents), and multimodal infotainment. Automakers and Tier‑1 suppliers are combining edge AI vision stacks for real‑time perception with cloud‑hosted AI data platforms for model training and lifecycle management, while governance tooling and marketplaces help meet safety, explainability, and deployment needs. Key platform categories include Edge AI Vision Platforms (real‑time object detection, sensor fusion, and on‑device model execution), AI Data Platforms (labeling, simulation, fleet data ingestion, and continuous retraining), AI Governance Tools (safety validation, model lineage, privacy controls, and regulatory reporting), Agent Frameworks (multi‑agent orchestration, goal management, and human‑in‑the‑loop controls), and AI Tool Marketplaces (model/component discovery, commercial licensing and OTA deployment packages). Representative tools illustrate these functions: IBM watsonx Assistant for enterprise virtual agents and multi‑agent orchestrations; Google Gemini as a multimodal model family and developer APIs for conversational and perception‑aware features; Anthropic’s Claude family for conversational assistants and analysis workflows; Yellow.ai and Lindy as platforms for building autonomous, human‑like agents with varying no‑/low‑code approaches; and Adept’s agentic research (ACT‑1) for automating multistep software interactions. As of 2026, the mix of stronger on‑device compute, multimodal LLMs, stricter safety/regulatory expectations, and demand for OTA extensibility makes careful selection of vision stacks, data infrastructure, governance, and agent frameworks critical for deploying reliable in‑vehicle AI.