YC is backing startups that, if stitched together, could start and run startups (eventually).
I know it is a bit far-fetched, but you’ll see it is a fun thought experiment that may help some founders identify startup ideas (gaps that YC startups aren’t filling yet).
In my last posts, I explored whether AI could take on the role of a founder and start a startup. I came to the (obvious) conclusion that it is still early days, but that a human-controlled AI could be useful as a founder assistant or, at best, a background co-founder.
I decided to see how far we would get by looking at the last YC Batch to create and run a startup with AI.
YC S23 AI Startup Landscape 👇
Introducing The Frankenstein Founder 👇
The Frankenstein Founder
I listed 7 skills across idea generation, startup execution, and leadership to see how far we get with stitching together a founder using YC AI startups.
✅ Problem discovery and creative thinking
Anari comes the closest. They are focused on exploring product ideas, refining business decisions, and saving time drafting docs all integrated with tools like Notion, Drive, Jira, etc.
Another interesting product in this domain is Info Bot which creates a news channel on any topic. An AI monitoring the news is an interesting way of identifying potential problems or opportunities.
✅ Idea validation
Although no startup is specifically focused on helping with idea validation for startup ideas, there are a few tools that can help with the process.
Raz dynamically creates forms to collect deeper insights from users. Roundtable on the other hand goes a step further to simulate surveys What makes Roundtable special is that you can use your customer data for training the model.
Beyond a dynamic survey builder and simulated survey responses, Outset provides insights through AI-moderated interviews to help you test your thesis and get user feedback.
LampBuilder (not YC) is an AI landing page generator that is great for testing startup ideas. Custom domain name hosting is free.
✅ Data analysis
Latentspace is an AI data assistant that helps to analyze data, create graphs, and do statistical analysis. Vizly leverages AI to transform your questions into actionable insights. Sensei is an AI engine that quickly identifies the underlying factors affecting key performance metrics.
✅ Customer understanding
Egress leverages data to help with customer-facing teams. On their website, they highlight product-led growth as a key use case. The value proposition overlaps a bit with customer support in the next section.
❌ Resource estimation
I imagine this type of service should evaluate what resources are needed by looking at market-related salaries, drafting a financial plan, cross-referencing it with third-party quotes, etc. I did not identify any YC startup focusing explicitly on this type of estimation. Perhaps a stepping stone for this type of service is what Flynt is doing to help finance teams bring accounting, reconciliation, ERP, and invoices together.
Although not specifically a tool to do resource estimates, Quill is a step in an interesting direction as a financial research assistant in financial markets. The demo on their website has screenshots of how it can search SEC filings and earnings calls, screen stocks, search for stocks matching your criteria using natural language, generate investment memos, and search through datarooms.
✅ Risk assessment
Alta is probably the closest example to help with risk assessment by summarizing legal docs and answering legal questions. YC also backed a few startups that focus domain domain-specific compliance. Greenlite is augmenting KYC, AML, and risk staff. HiddenHand is a tool for law firms to screen clients, reduce response time, and cut onboarding costs.
⏳ User experience design
I imagined a service that can create mock-ups and user journeys to fine-tune the ultimate product user experience.
The closest type of service I could identify is Watto which acts as a companion writer for Product Managers and AutoEmber which creates websites without code.
✅ Develop MVP
For me, Spine AI is probably one of the most interesting YC startups in this batch. They turn any product into a conversational interface, which I believe will be a great way to create MVPs.
YC backed quite a few startups that are focused on software development. Metoro helps debug production issues faster. Quack AI helps onboard new developers. Ten.dev is a tool that automates the creation of detailed task specifications and code reviews for developers, making it easier for them to focus on coding. Codestory is a new code editor using AI to generate code, execute terminal commands, refactor, fix bugs, review PRs, and handle git, too. Continue.dev is similar but as an extension for VS Code. Sweep.dev specifically works on your GitHub backlog in the background.
The early developer tools are designed with having a human in the loop. It will be interesting to see how this evolves. Axilla is an interesting example. They have created an opinionated LLM framework for TypeScript engineers of the future: a product-focused TypeScript developer.
Outside of software development, Osium does Machine Learning for Materials Science to speed up materials innovation.
⏳Quality assurance
I expected to see more quality assurance tools for making sure the code works as intended.
Structured Labs helps with log analysis and Remy Security does AI-powered security design reviews. Talc helps with the quality assurance of LLMs by running your AI against hundreds of realistic use cases so you can ship with confidence.
❌ Project management
I have not seen a specific project management AI startup.
Perhaps the closest is Podtrace, an AI-driven platform that integrates your BOM with your supply chain, and Asklio which is working on automating procurement.
✅ Financial management
Fynt leverages AI to bring accounting, reconciliation, ERP, and invoices together.
Respaid is focused on collecting outstanding payments. Similarly, Subset is helping with retaining subscriptions by enabling personalized and proven retention initiatives.
Slightly unrelated services are Cascading and Onnix, which help bankers with their internal workflows like processes like onboarding, customer service, and back-office automation.
✅ Customer support
YC backed quite a few customer support AI startups.
Octo uses AI Agents to resolve support tickets instantly.
Hyperlight uses enterprise-grade AI Agents for customer support.
Poozle is an open-source unified API for ticketing.
Rightpage is an internal AI tool that provides detailed answers based on product documentation.
❌ Marketing
I did not find any startups using AI to do marketing.
✅ Sales outreach
Everything from finding leads to closing deals!
Inventive automatically responds to sales information requests with AI such as RFPs, security reviews, DDQs & RFIs.
Humanlike is a cloud platform for deploying voice bots with human-like intelligence.
Hyperbound adapts automated emails by researching prospects using domain-specific knowledge in your CRM and public internet data sources to generate high-quality, reliable personalized emails at scale.
Docsum helps to close deals faster with the power of AI through contracting workflows from review to negotiation by 3x.
❌ Recruiting
I did not find any startups using AI to do recruiting.
✅ Legal and compliance
YC backed quite a few startups focusing on legal tasks:
Safety Kit offers AI-driven policy enforcement that allows for immediate decision-making based on predefined rules, eliminating the need for costly training and providing transparent reasoning for each decision.
Case Hopper is an example of domain-specific AI services for writing immigration documents.
Agentive is an AI-powered auditing service. You can ask questions, perform audit procedures, and generate documentation.
Sweetspot specializes in sourcing and placing bids on government contracts that best match your company's services.
Dili is a diligence platform built for Private Equity and Venture Capital deal teams to make more data-driven investment decisions. Dili applies AI across 3 pools of data.
Solve Intelligence and write patents with AI.
⏳ Communication
I expected to see tools that integrate directly with Slack and can stand in as a team member to either answer questions, provide links to previous conversations, etc. A step beyond communication would be to show which team members work together effectively, where there is a delay in actioning the next steps, etc. That being said I did find two startups working on communication tools:
Onward makes it easy to automate your most important conversations.
Sidenote is a Chrome extension that turns notes into action items, drafts emails, creates calendar invites, Jira issues, and more.
⏳Decision making
For AI to step in as a founder, it would need to be able to make all kinds of decisions: direction, what to prioritize, who to hire, etc. Although there aren’t any startups that focus on this type of decision-making explicitly, it is interesting to see Pincites that specializes in closing deals faster with AI for contract negotiation.
❌ Team building
I did not find any startups using AI in team building.
❌ Performance evaluation
I did not find any startups using AI for performance evaluation.
❌ Conflict resolution
I did not find any startups using AI for conflict resolution.
❌ Emotional intelligence
I did not find any startups using AI to enhance emotional intelligence.
🤖 Other
There were a few startups that I thought could also fit into the picture of stitching together an AI founder. Sola Solutions and ReworkdAI help to automate repetitive and manual workflows.
Your Nowadays is an AI-powered event planner that takes the hassle out of organizing corporate events. Just input event details and our AI will contact venues and deal with negotiations for you.
Linc instantly parses documents and text from your emails and syncs this data with your system of record.
Guac does demand forecasting for grocers.
Conclusion
The latest YC batch reveals a tantalizing possibility: a "Frankenstein Founder" assembled from various AI startups backed by YC. While AI excels in areas like data analysis and financial management, it falls short in emotional intelligence and team building.
A lot of the skills are not specific to a founder, but rather running a startup. So, it looks like we're heading towards a future where human founders and AI work hand-in-hand. Bottom line: The Frankenstein Founder isn't here to take your job, but to make running a startup a whole lot easier.
I found an AI to read your post