Each year Y Combinator release their RFS (Request for Startups) idea document for founders as a guidance on what they would like to favor during funding. They just released their 2024 list, and it looked very interesting. So I decided to write a brief post covering their/my opinion on some idea presented on the list.
Applying machine learning to robotics
So, GPT moment for robotics not happened yet. However, with recent advancement in foundation models, it should be now possible to have nearly human level perception in robots. As they say there's a big market potential for hardaware/software tooling that help people build/optimize robots. I think mass use of next gen robots will happen initially in B2B market not consumer space.
Using machine learning to simulate the physical world
They are interested in startups that replace existing simulation software with ML-based ones. Lot of industrial simulation tools have potential to be replaced with better machine learning based tools. Those solutions have potential to be much less computationally intensive.
Climate technology
Huge potential ($3-$10 trillion market) in commercial solutions to decarbonize society and remove carbon from atmosphere. Most regulation in the world will also help to accelerate existing market trends. There's a once in a generation opportunity up for grabs here.
Commercial open source companies
Almost all closed source big enterprise software sucks. They move slowly, adapt slowly and slow to use. Open source companies can move more quickly than closed source companies and can become mature quikcly to sell to enterprise. Very technical founders have strong advantage in the space to drive sales engine purely on technical merit and iterate faster when going in open source path.
Spatial computing
With the recent release of Apple Vision there's huge market space to build spacial computing apps with better UI/UX. There's a market to solve practical real world use cases beyond just gaming.
New enterprise resource planing software
ERPs are known to be expensive, painful to implement and disliked by users yet necessary to business. They are interested in products that help to fulfill this gaps by building products that's flexible and loved by.
Explainable AI
Understanding current large AI model behavior is challenging. It's critical for AI safely/trust to build better and trustworthy tools to explain AI model behaviors. More work needs to be done on building tools to explain AI.
LLM tooling for back office processes in legacy enterprises
In almost every large/old company there are huge teams of people running manual processes. Those tasks were very difficulty to automate before existence of LLMs. LLMs now allows whole categories of manual processes to be automated in ways that weren't possible until recently. Huge opportunities in the market to build tools for back office automation with a new LLM perspective.
AI to extend enterprise software
Enterprise software has reputation among developers that they are being boring to work on. You build something and a potential customer want something slightly different. You ended up building bloated software to please them all. What if AI could write those slight modifications for them. You build a solid generic core first. You or client can safely extend the core using AI. AI is already reasonably good at extending existing code bases. In future each enterprise could have their own custom ERP, CRM or HRIS that is continually updating itself as the company evolving. This is my favorite idea of this year!
Stablecoin finance
It seems that stablecoins will play a big part of the future of money. They can utilize blockchain or not, the priority is solving real world problems. They are inviting founders to build consumer products on top of stablecoins, tools and platforms to enable better stablecoin finance.
MSO (Managed Service Organization) model for healthcare
A new startup model has emerged as an alternative to private equity ownership. The MSO (Managed Service Organizations) model.
MSO model can enable healthcare professional run their own clinics, while back office tasks handled by a software service. In my opinion this model can be applying to whole range of industries to empower consumers to start their operations without high equity requirements.
This was really well-thought-out list this year. YC funded or not we are eager to support founders who are eager to tackle any small/large problems here at Xaventra. Ciao!