DevConnect is less than two weeks away, and with it comes Base Batches Demo Day on Nov 19th. Out of 50 finalists competing for funding and ecosystem support, 7 are building at the intersection of AI and crypto—from yield optimization to onchain gaming to tokenized APIs. Here's your guide to tracking these teams, gaining edge before the market catches on👇 ~~ Analysis by @davewardonline ~~ For context, Base Batches is @base's accelerator-esque program for teams building across two tracks: the Builder track for pre-product teams and the Startup track for teams with live products. The program's been running since early September and is now rapidly approaching its finale on Nov 19th, where Startup track finalists will demo in front of angel investors, VCs, and accelerators for the chance to receive funding and ecosystem support. Out of the total 50 finalists, 7 of them incorporate or build for AI, targeting use cases from yield to APIs to dating. Below I've profiled these teams to give you a guide to tracking them and hopefully gaining market edge in the process. @0xNeurobro is a Virtuals-based, AI-powered intelligence platform for tracking and finding alpha. Built for surfacing actionable insights like smart money token accumulation, it maintains a database of both whales and tokens, showing buy/sell volume, latest activity, and realized PnL. Delta is an AI built by mathematicians, quants, and former hedge-fund ML engineers, designed to capture market-neutral opportunities across DeFi. Backed by grants from Consensys Fellowship, Uniswap Foundation, and Arbitrum, Delta's Dune dashboards show an ROI of 98% since October 7th. It works through Delta-neutral market making on DEXs, yield farming, and other complex DeFi strategies. @codygame_com is an agentic gaming platform aiming to build the first $1B onchain game through social, casual onchain gaming. Playable in both the Base app and Farcaster, CODY already has a token that serves as fuel for game entries, in-game purchases, and rewards. @NetMindAI is a fullstack solution for decentralized AI, offering plug-and-play API services such as model integration, as well as an MCP hub for integrating with existing services. The platform also provides GPU infrastructure and business solutions. @apinowfun is building on x402, creating tokenized APIs that let AI agents directly purchase access using tokens. Agents can autonomously discover relevant APIs while token holders gain fractional ownership of these flows. Each call burns tokens, creating a deflationary mechanism tied to actual usage. @tryEmerge is an AI mini-app for monetizing onchain workflows. Users create content workflows by tagging the Emerge agent on Farcaster with aesthetic direction, which then generates a mini-app that others can pay to use. Creators earn rev share on each use, and the team is planning an airdrop post-TGE. Call Me Sensei is an onchain AI agent dating game teaching Japanese, available via iOS and Google Play. The concept blends AI agents, language learning, and dating sim mechanics. These seven projects showcase the breadth of what's being built at the intersection of AI and crypto on Base. For those looking to gain edge or discover new projects early, keeping tabs on how these teams perform at Demo Day could prove worthwhile. When winners receive funding and ecosystem support, their tokens can take off, potentially giving early followers a head start on the market. It may be good to set aside some time on the 19th.