23 startup ideas — the Next Play Request for Startups March 2025
Next Play Request for Startups March 2025
Here’s a common situation amongst people in tech: they have skills and a work ethic — but they have no idea what to work on. And so they explore for quite a while and in that process start asking around: “what should I work on?”
We often get messages from people in the community wondering if we have any startup ideas or problems that we would like solved. And while we ourselves do not have many great answers to this question - we decided to ask around and crowdsource interesting answers from people in the community (mainly venture capitalists and experienced operators).
We compiled their answers into a list that contains a range of ideas - everything from a “digital AI twin for your loved ones” to “better dating apps.” They should give you inspiration around what to work on (and you may even want to reach out to the people who shared the ideas if you end up building them).
Here is the “Next Play Request for Startups” (March 2025):
Amir Shevat: Please build a solution to replace Workday. It sucks and many managers complain every time they go to that system.
Michael Stoppelman: A service that creates an AI digital twin of your love ones (specifically parents). I'd like an ai that can interview a loved one and create a digital twin for answering questions after they pass. This is meant to be for grandchildern and other descents to ask questions about their life.
Terrence Rohan: Not quite a startup, but an Agent driven 'Autonomous' VC fund. Founder chats with an Agent, uploads materials and an investment decision within 1 hour; agent reviews docs and wires same day (within 24 hours). Writes smaller non-lead checks (i.e $1M of $3m round). Army of agents focus on "value-add" for the company. Founder friendly, infinite scale - complementary to existing VC ecosystem (for now;).
Sean Whitney: There should be AI enabled travel agency for golf trips. Travel golf is a massive and growing global market. Outside the US, a number of golf tour operators exists as low margin services businesses. You need some tech on the backend to integrate with hospitality management systems and tee sheet management systems. But a lot of work done by the folks in the call center during initial touchpoints should just be done real time with AI in a new and inviting digital UX.
Alex Brussell: A dating app where you get only one swipe called One In The Chamber.
Camille Ricketts: Better AI scheduling assistance. In particular, if it could also make appointments and reservations and place orders reliably for you. I think it's time we should have personal agents we talk to all day long that should be capable of these more complex operations in accordance with our voices and preferences.
Camille Ricketts: Software for healthcare providers that allow them to save time and money and provide better, more present care for patients.
Camille Ricketts: Vertical AI applications for dusty and tech-forgotten industries within defense, aerospace, energy, unsexy backend or back office software (Quickbooks, Netsuite, etc.), and physical manufacturing/supply chain.
Aashay Sanghvi: A consumer marketplace for real estate services. I'm interested in the model of companies like Manifest Law, which aggregates demand for immigration and employment legal concerns and scales the supply of lawyers through AI. I wonder how this model could be applied to serving consumers who want help with discrete real estate tasks -- anything from looking over rental agreements to purchasing homes sold in auction.
Aashay Sanghvi: AI-native fraud and risk platform. As agents crawl and roam the web with heightened speed and intensity, the need for fraud, risk, and identity providers that can delineate between human and non-humans increases. Previous platforms were architected for a web 2.0 era and use blunt force against non-human bots. I'm also interested in how AI-native workflows can speed up the operations of teams inside e-commerce merchants and fintechs as they navigate this new landscape.
Lan Xuezhao: Infrastructure-first consumer companies. Read more here.
Lan Xuezhao: Industrial robotics companies - anyone working on non-consumer robots.
Tanay Jaipuria: AI Inside Sales Agent: We've seen a lot of AI SDRs and other similar products, but I'd love to see startups take a crack at building something akin to an inside sales rep that perhaps sits on a website and converts prospects to customers. This wouldn't be a chatbot, it would likely be an audio or video powered sales rep that can engage visitors, understand their needs, given them personalized demos, handle objections and potentially negotiate and also do the final end contracting via email. I can see something like this being applicable in both B2C settings for purchases where reps don't make economic sense today or for B2B companies that perhaps have ACVs that make it difficult to efficiently scale out their team or just have terrible website traffic to sales conversion rates.
Nicole Wischoff: AI agent to underwrite a mortgage or even do a partially underwritten pre-approval in less than 24 hours. B2B play, selling into existing lenders.
Nitesh Banta: There should be an AI tutor for early childhood education. AI's impact on education will be profound. I'm excited about new modalities to personalize education at the earliest stages when many parents are unsure how to introduce technology to infants and toddlers.
Adrian Nicholas Radu: Infrastructure for Agentic Finance. We've seen financial agents handle a number of capabilities across personal cash flow, treasury management, tax optimization, and portfolio reallocation etc. We're in the early innings of DeFi applications with smart contract automation (yield, lending, etc.) and will see more composable financial agents that execute across systems.
Adrian Nicholas Radu: AI-native PAS (Policy Administration System). We've seen a number of attack vectors across underwriting copilots, claims automation replace and/or supplement traditional PAS functionalities. There is an opportunity to redefine the core system across the policy lifecycle from issuance, underwriting, billing, claims etc.