Nika

Y Combinator offers 7 startups ideas they want to fund (Spring 2026)

As usual, Y Combinator came up with segments that are worth investing:

1. Cursor for Product Managers
2. AI-Native Hedge Funds
3. AI-Native Agencies
4. Stablecoin Financial Services
5. AI for Government
6. Modern Metal Mills
7. AI Guidance for Physical Work 
8. Large Spatial Models 
9. Infra for Government Fraud Hunters 
10. Make LLMs Easy to Train

Which of the following do you think are the most interesting?

OR

Would you add to the list, maybe something that is more worth funding?

Also, are you going to apply? 👀

2.9K views

Add a comment

Replies

Best
gregb

What about an AI-native investment bank or fundraising broker?

A few players such as Metal, DeepFlows, Signal are heading towards this positioning helping startups raise money against a subscription at first, and eventually taking a success fee out of the deal. What do you think?

Nika

@gregbtb It sounds fancy. How it should work according to you?

gregb

@busmark_w_nika the platform handles matching, warm intros, outreach, tracking. Only a small, curated subset gets human advisory on top.

Model:

-Platform users = subscription

-Assisted startups = success fee (outcome based)


This model helps both ideation/pre-seed startups in discovery, matching with incubators/early investors, as well as VC-backed or incubator-backed startups with their next rounds

Nika

@gregbtb This idea doesn't sound bad. The thing is, who will make it happen? Are you about to create something like that? :)

Fabio Salvadori

Agentic AI safety and security are what will make all the sectors in the list investable, and we are really behind at the moment, so for me that's the number one sector worth funding.

Nika

@fabiosalvadori I agree with this one. Just let's have a look at how soft something is without knowledge, when it is vibecoded. It will break in a millisecond. Data leaks guaranteed. Are you building something in this sphere?

Fabio Salvadori

@busmark_w_nika Thanks Nika. Yes, I launched yesterday, but with problems and it got buried: https://www.producthunt.com/products/pic-standard-causal-firewall
Not a big deal...PH is not best place for open source dev tools anyway :)

Nika

@fabiosalvadori Sorry to hear that. Maybe relaunching on other platforms could help with link building and partial visibility. :)

Suzanne Chartier

@busmark_w_nika #1 really resonates — Cursor for Product Managers is exactly where we’re seeing the biggest gap right now.

At Agiloop, we’re building an AI-native product development loop that starts with guided intent capture (persona interviews → clear stories, scope, estimates, and outcomes) before anything gets built — not just faster docs, but better decisions.

AI is crushing execution speed, but what teams struggle with now is what to build, why, and how to continuously adapt once real usage shows up. PMs need AI copilots for thinking, not just writing tickets.

Feels like this category is going to define the next generation of product tools.

Nika

@suzanne_chartier Have you launched it yet? Or are you about to polish and launch? :)

Suzanne Chartier

@busmark_w_nika We launched on Product Hunt earlier this week. The INVENT product has been available since early January.

Shaunak Sayta

Hi! The list is very interesting, I have applied for Spring of 2026 and I think my startup fits in the AI Guidance for Physical Work category. I'm working on an AI tool to intended to help you take steps to improve tree health, and keep track of tree health, so as to achieve better outcomes, yield improvement (in case of fruit trees), and climate resilience (in some cases), better tolerance to heat and stress (mitigate climate change). Essentially like Google Analytics for trees so we have that continuous feedback loop from the tree in the form of images, and soil data so repeated interventions can help longer term better outcomes.

Nika

@shaunak_sayta The question is: How do you want to monetise this business?

Shaunak Sayta
@busmark_w_nika We have to charge a license fee per tree - for orchards. The incremental yield and reduced fertilizer cost will justify it.
Girik Gangwani

Isn't " @Cursor for Product Manager" already available - as @Claude Cowork?

Nika

@Claude  @girik_7 Isn't Claude cowork for more general tasks? At least I perceive it that way.

Girik Gangwani

@Claude  @busmark_w_nika It was, until they launched the 11 plugins on 30th Jan, which essentially are those field's specific expertise tools that caused software native companies' shares to tank, all around the globe!

Nika

@Claude  @girik_7 In that case, some closer clarification by YC could be cool because at this point I do now know what "innovative" they want to see in these terms.

Filip Holec

They seem to have updated the article and I really like Infra for Government Fraud Hunters - even though mainly focused on the US, pretty sure that in EU, especially central / eastern countries the fraud and corruption is quite high.

Also really like the push to make the LLMs easier to train!

Nika

@fholec Thank you for the update, yeah, they added a few more items, gonna add them to the list :)

Rob Maclean

#1 is interesting. But it needs to address how the PM role itself needs to change with AI. Scrum-based planning misses the new super-power - multi-phase planning. A chat-based tool that can create PRDs (and keep them alive by managing change requests the same way we manage Git) in a structured data format JSON) that eg Cursor but platform-agnostic can read via MCP to formulate its own plan (asking clarifying questions, approval gates for plans etc) would be the game changer. AI Jira isn't exciting (Ask Rovo). PMs need to move upstream from scrum micro-planning to focus on the requirements layer - Requirements As A Service will be an entirely new service category. That would be interesting.

Nika

@rob_maclean2 If AI can work with complexities (yes, it can), it would improve the whole process, of course... the thing is whether teams can work at such a pace. I still somehow feel the human nature to be prepared for huge changes.

Princewill Okafor

Oh no wonder i got rejected from YCombinator😩 ...... My product isn't in this list.

My Vote for the list would be [2], [5], [4], [6]

I would add AI to create assessment from Learning materials & video

I dont understand the other product category

Nika

@oprincew What was your product you tried to apply with?

Renato Marinho

#1 and #3 hit different for me.

Cursor for PMs is wild because most PM tools still feel like 2015 — imagine actually managing sprints, standups, and priorities through AI right in your IDE. That's where we're headed with GitScrum MCP Server (literally why we built it).

AI-Native Agencies is the sleeper hit. Agencies are drowning in client work, and autonomous agents could flip the whole model — less burnout, better margins, faster delivery.

What I'd add: AI-native dev tools that talk to your entire stack — not just code, but PM, CRM, time tracking, all through one protocol. MCP is early proof of concept.

Not applying yet, but watching closely 👀 What about you? Building something in one of these spaces?