Ideas Radar: May 27, 2026

Founders keep describing the same hole in B2B SaaS: a real customer-communication product. Not just another transactional email tool, but one that unifies every channel (email, SMS, in-app, push) and lets the end user pick their preferred method, with an inbox-style notification center like every social platform already has. The pain is that companies bolt together five tools to fake this. A single product that owns the user's communication preferences and routes everything through them is a clear willing-to-pay wedge.
Source: https://x.com/DanBochman/status/2058868173730963610

As agentic shopping arrives, one builder spotted the asymmetry everyone is ignoring. Google and others are racing to build agents that buy your necessities (pens, eggs, refills) where no browsing is needed. But luxury and considered purchases, beauty, perfume, handbags, convert under 0.5% online versus roughly 10x that with a human salesperson, and nobody is building the agent for that side. The opportunity is an AI personal-shopper/concierge that recreates the high-touch salesperson experience for high-margin goods, exactly where the money and the unsolved conversion problem are.
Source: https://x.com/Jazz_from_Korea/status/2058842426769658147

A small, concrete consumer ask with surprising reach: a phone app that constantly shows the live price of an asset (BTC in this case) right in the status bar and on the lock screen, refreshing every 15 seconds. The person looked and genuinely couldn't find one. It's tiny, but "glanceable live data permanently pinned to your lock screen" generalizes far beyond crypto, stocks, sports scores, package tracking, server status. A simple, ambient-data utility with an obvious paying audience.
Source: https://x.com/HermesLux/status/2058961320796471298

A developer wish that doubles as a real service: an iPhone simulator running a US App Store, on demand. Lots of people outside the US (or testing geo-locked apps) need to see, install, and trial App Store apps as if they were American, without a US Apple ID or a physical US device. A hosted "US iPhone in the cloud" for app discovery, QA and competitive research is a niche but genuinely under-served tool.
Source: https://x.com/upbeat_dev/status/2059018290907709614

A sharp hardware-plus-data idea aimed at the robotics gold rush: a company that ships you a pair of two-camera glasses to record your everyday activities in ego-centric stereo, so that footage becomes training data for humanoid robots, with you contributing from home (sorting LEGO, tidying after kids, normal chores). Everyone needs embodied training data and nobody has a clean consumer pipeline for crowdsourcing it. The product is the glasses plus the marketplace that pays people for their mundane footage.
Source: https://x.com/Cocoanetics/status/2058912482660425992

A recurring complaint from people closer to the SMB world: models keep getting smarter but it hasn't translated into businesses actually automating, because most companies want to automate workflows yet don't know how or can't scale it fast enough. The gap isn't a smarter model, it's a productized, vertical, done-for-you automation layer that meets non-technical businesses where they are. Whoever turns "I want to automate this but can't" into a packaged, deployable service for a specific industry captures a huge underserved middle.
Source: https://x.com/Jamthelion/status/2059005568619086306

A niche developer-tooling gap worth noting because niches like this are exactly where indie tools win: native Vim mode in the Arduino IDE. The Arduino ecosystem is enormous and growing, the official IDE has no proper modal editing, and the people most likely to use it (embedded tinkerers) overlap heavily with Vim users. A small plugin or fork solves a real daily annoyance for a large, specific community.
Source: https://x.com/jerkeyray/status/2058863459614535868