Y Combinator Startups in New York that are currently hiring 2025

We’re a passionate team of dreamers and builders, determined to liberate small businesses from the not-so-fun aspects of the job so that they can focus on what’s fun and important.

SmartAsset is an online destination for consumer-focused financial information and advice. Reaching approximately 59 million people each month (as of January 2024) through its educational content and personalized calculators and tools, SmartAsset's mission is to help people make smart financial decisions. Additionally, SmartAsset powers SmartAsset AMP, a national marketplace connecting consumers to financial advisors. Building on SmartAsset’s relationship with millions of investors, SmartAsset AMP connects financial advisors directly with prospects who meet their target client profile. Our algorithm matches consumers with up to three fiduciary financial advisors. In 2023, SmartAsset helped advisors close an estimated $34 billion.

F2 is the AI platform for private markets investors. Purpose-built for private credit, private equity, and commercial banking, F2 accelerates diligence, underwriting, and portfolio monitoring, helping teams evaluate deals over 60% faster. F2 turns unstructured deal data into investment-grade analysis and polished memos in minutes. Unlike general-purpose AI, F2 is built with the domain expertise and depth to handle the most complex financial analysis. Already trusted across thousands of deals, F2 makes collaboration easy and retains historical context, so past work compounds into future wins. Headquartered in New York City, F2 is backed by leading investors including NFX, Left Lane Capital, and Y Combinator.

The Interface enables anyone to create, interact, and share on the fly — from TV shows to games to movies. Drawing on deep expertise in realtime compute and simulation, we combine multimodal AI with the best of 3D simulation and modelling techniques, to deliver real-time experiences that are alive, interactive, and emergent.

Prosper builds AI phone agents for healthcare providers. Our agents automate manual, time-intensive workflows like patient scheduling, balance reminders, claims status checks, benefit verification, and prior authorizations. We integrate with 80+ EHRs, go live in 3 weeks, and offer a no-code interface for customization. Today, Prosper handles thousands of daily calls for dozens of providers, reducing costs by 50% and tripling productivity. We’re building the AI-first operating system for healthcare to eliminate the $200B lost to administrative overhead every year.

Fleetline is building the first complete-context algorithmic load planner for mid to large-sized trucking fleets. Today’s dispatchers face an overwhelming task: balancing regulations, customer demands, live fleet data, and individual driver needs, all while coordinating across siloed tools and teams. Mistakes run rampant, and each mistake is thousands of dollars. Even the select fleets using algorithms that have fleet-wide context struggle: the outputs are rigid, hard to interpret, and are blind to real-world nuances like driver preferences and schedule exceptions. Fleetline solves this by combining advanced optimization with LLMs that can easily adapt algorithms to each fleet’s needs and capture driver-specific information. The result is smarter planning truly optimized for every fleet and its drivers.

CareSwift is a real-time AI assistant built directly into ambulance reporting systems, guiding medics as they document care and instantly flagging errors. Crews complete accurate, compliant reports in under three minutes, cutting documentation time by 80 percent and helping agencies avoid insurance claim denials. Ambulance crews lose valuable time completing reports after each call. Even minor mistakes cause insurance claims to be rejected.

Teach robots new skills through natural language and run reliably in production in minutes. Mbodi combines generative AI and agent orchestration, turning everyday language and quick demos into precise, reliable robot actions for industrial automation.

Acolite builds AI teammates that handle submissions, issuing COIs, and SOV mapping across any AMS. Like an extra teammate, you simply forward an email and Acolite handles the rest.

Law enforcement is drowning in data. Closure helps search it to solve crime. - Hiring: UX, Engineering; - Expanding: Local, state, and federal law enforcement.

Trata is an AI-powered research desk for hedge funds. We use agents to interview hedge fund analysts, write investment research, and distribute this through a subscription.

The student information system (SIS) is the core system of record in every K-12 school. It tracks students, attendance, grades, schedules, staff, and more - kind of like the ERP for schools. Legacy SIS providers do just about that, and not much more. School districts layer dozens of tools on top of their SIS to fill in the gaps. Scout is an AI-powered student information system that automates school- and district-wide busywork, from day-to-day operations like attendance tracking to annual compliance reports. Our mission is to help school districts run like high-performing organizations, enabling them to provide high-quality, personalized instruction to every student.

Karsa is a stablecoin neobank for over 100M people in emerging markets. We enable people to buy dollar-denominated stablecoins, save money in US bank accounts, and spend anywhere with a Visa card.

Misprint is an all-in-one marketplace that solves the biggest pain point in the collectibles world: opaque pricing. Collectors, investors, and hobbyists currently rely on scattered data from auction houses, online forums, and reseller platforms, making it difficult to know the true value of their items. Misprint streamlines the process by combining accurate pricing analytics with a seamless buy/sell marketplace, making it easier than ever to understand and monetize collectible assets. Existing marketplaces lack robust, real-time analytics because accurate valuations require deep expertise in both the collectibles domain and advanced pricing algorithms. Misprint’s unique blend of insider trading card knowledge and machine learning allows us to fill this gap. Building these pricing models requires extensive data infrastructure, deep domain expertise, and cutting-edge ML. Recent surges in the trading card market show that collectors crave transparent valuations. The global collectibles market is worth tens of billions of dollars. Eva has experienced the scale of this industry firsthand, generating $500k ARR within three months of taking her Pokemon card-selling side hustle full-time. With additional segments like sports cards, comics, and emerging collectible categories, the growth runway is immense. Co-founders Eva and Jon have been best friends since Exeter, combining finance expertise, a passion for collectibles, and mathematical rigor. By merging Goldman Sachs-level financial analytics with top-tier mathematical talent, Misprint is uniquely positioned to become the go-to platform for fair, transparent pricing in the collectibles space.

Keye is an AI software that lets private equity firms do due diligence for acquisitions faster and more accurately. Our customers can evaluate 10x more deals and fund better companies which leads directly to better returns/more alpha. -- Our system is the first of its kind to instantly generate complex, comprehensive data packs from raw data in any format, directly from the data room. This enables investors to gain an 80% understanding of a potential deal within moments, empowering them to say No to a deal earlier in the process and focus on high-potential opportunities. Traditionally, creating these detailed data packs—like cohort or retention analyses and customer roll-forwards—takes weeks, during which critical details may be overlooked, potentially costing funds millions in missed insights and poor decisions. -- Keye combines cutting-edge AI with deep private equity expertise to deliver powerful, multi-layered insights that go far beyond the simple summarizers and search functions offered by other tools. Our platform dives deep into both quantitative and qualitative due diligence data, offering high-impact analytics crafted specifically for private equity workflows. This isn’t just automation—it’s a purpose-built solution that uncovers actionable insights, fueling smarter decisions and driving true alpha.

Asha Health is building the future of self-learning generative AI for healthcare.

Afternoon.co empowers founders to make better business decisions with real time financial metrics. We pair managed AI bookkeeping and tax services, to reduce administrative burden on founders. Accountants struggle serving 4 million+ of ecommerce brands & startups in the US, because of complexity in sales tax, multiple sales channels and inventory management. We automate a large portion of bookkeeping by integrating with their sales and inventory systems, so founders can spend less time doing admin work.

Symphony helps enterprises train their customer-facing employees using interactive AI voice simulations of real-world scenarios. Users can create an infinite number of simulations that allow their teams to practice the job - avoiding costly mistakes with real customers. For example, we’re live with 2 of the top 3 auto insurers in NA, helping train 1000s of newly hired claims adjusters.

Raycaster is regulatory intelligence for biopharma documents - think Cursor for regulated content, helping medical writers and CMC teams draft, edit, and review submissions much faster.

https://trypear.ai

Historically, quantitative models are domain specific. Brilliant people spend their best years testing features, tuning hyperparameters, and iterating architectures within a narrow domain. But scale is the panacea: large models will find patterns people, and specialized models, could not. Forecasting generalizes. Zoa trains cross-domain event forecasting engines. *Automating Iteration* LLMs—embedded in multi-agent optimization loops and evaluated against fixed policies—can automate the build-test-improve modeling cycle. Think AlphaEvolve for forecasting problems. *Sample-Efficient General Models* Today’s forecasting models are narrowly crafted with deep human priors. But larger models will outperform state-of-the-art specialized models. Unlike existing event models, our models leverage data from across contexts and rely less on human intuition. And compared to LLMs, our models are built with more inductive priors and rely more heavily on inference-time compute—improving sample efficiency. *Why It Matters* In the real economy, our models could be useful for forecasting supply chain volatility, energy supply and demand, even earthquake risk. Science is, Ian Hacking writes, the taming of chance. It is the process of iteratively updating priors (something like: identify uncertainty, conceive experiment to reduce uncertainty, execute, update). If science is uncertainty-reduction, forecasting is a critical measure of progress. Better forecasting improves our ability to select interesting experiments (roughly those with greatest expected uncertainty reduction) and update priors. Our models will be used by labs and academics in data-heavy domains. Sam's ex-girlfriend introduced him to Greg back at Carnegie Mellon in 2017, and while that relationship didn't last, their friendship has. After college, Greg went to Harvard Law School, while Sam worked for three years at Jane Street on their Options desk, building & leading a satellite dev team.

Ultra builds practical, general-purpose robots that are deploying—and producing ROI—today. Unlike traditional industrial automation that’s rigid and complex, our robots are zero integration (able to be installed in hours, not weeks) and are highly flexible, capable of quickly learning new tasks and delivering immediate value to customers. Founded by a team of three-time entrepreneurs with a decade of collaboration, Ultra moves fast. We already have robots in the field generating revenue and data, with plans to rapidly scale deployments this year.

Diode automates circuit board design. We work with hardware companies to get them the PCBs they need, fast.

Codes automates patient record collection. Today, documents are scattered across EHRs, providers, and faxes; we use AI to compile & analyze docs for those who need them.

Elevate is bringing AI to private equity-backed roll-ups. For example, one of our customers is Guardian Restoration Partners, which is backed by Alpine Investors, the leading roll-up focused private equity firm. In three months, we integrated 10 acquisitions and enabled them to pilot AI agents built on top of our data warehouse. We know this problem personally. Before Elevate, Scott was a Vice President at a $15B private equity firm where he was leading a roll-up. Steve and Jeff are both software engineers with backgrounds across finance, big tech, and start-ups. The founding team has been friends for over a decade from their Duke undergraduate days.

The American healthcare system is complicated. 88% of US adults cite difficulties navigating the healthcare system. The good news is patient advocacy became a covered benefit by Medicare in 2024. Understood Care matches Medicare patients with a healthcare personal assistant (known as a Patient Advocate) at little to no cost to the patient. Patient Advocates help patients patients access care - finding resources, helping fill out paperwork, and much more. Understood Care uses AI and healthcare expertise to deliver the highest quality patient advocacy to the 68 million Medicare patients in the US.

OffDeal is an AI-native investment bank, offering small business owners a faster, cheaper way to sell their companies at premium prices. Our AI technology automates most of the grunt work typically done by analysts at larger Wall Street banks, enabling our in-house M&A advisors to focus more on strategic client interactions, leading to better outcomes for both sellers and buyers.

Clarion is building the most advanced AI communication layer for healthcare. Our AI agents handle the overwhelming volume of calls and messages that clinics face daily—from scheduling to billing to prescription refills—freeing staff to focus on patient care. We're an in-person company based in New York. Today, clinics miss 30-40% of inbound calls due to staffing shortages. We're solving this by automating complex workflows end-to-end, already serving tens of thousands of patients monthly across virtual care companies, health systems, and a $5B health insurance company. We've raised $5.4M from Accel, Y Combinator, Sequoia (scout), and healthcare founders from Ophelia, Medallion, and Counsel Health. Our founders bring both clinical and technical expertise—our CEO Ryan is a Stanford/Harvard-trained physician who was on the founding teams of Two Chairs and Ophelia, while our CTO Jeffrey built voice AI at Amazon Alexa and led AI/ML teams at Salesforce. Join us as we transform healthcare communication—creating a future where patients get faster resolutions and providers reclaim time for what matters most.

Hatchet abstracts away the infrastructure for managing task queues and message brokers, so you can focus on writing your background task and workflow logic. With Hatchet, you can run slow OpenAI requests in the background with async tasks, chain complex tasks together into workflows, and set retries and timeouts to recover from failure.

Garage is a marketplace for used equipment and vehicles, starting with government surplus. By automating every part of the transaction—including payments, freight, paperwork, warranty, and financing—we reduce what used to take months down to minutes. We're building the modern platform for equipment that powers America’s essential industries.

Understand & control neural syntax.

Yarn uses AI to help founders, salespeople, and marketers make product and sales videos. Teams are using Yarn to make sales collateral, launch videos, use-case demos, training guides, and feature updates for Twitter and LinkedIn. No video experience needed and everything looks great out-of-the-box – think Notion, Linear, or Stripe-quality – but branded for your team. The magic really comes from AI (TTS, CV, LLMs) – so you can zip through 5x faster than a Loom.

CrowdVolt is rebuilding the infrastructure for live events, starting with an automated marketplace for raves and electronic dance music.

Hazel builds AI-powered procurement software for U.S. state, local, and federal government agencies. We automate the full procurement process including requirements generation, market research, solicitation drafting, and proposal evaluation. Agencies use Hazel to save time and money while improving mission delivery.

Lawyers have to manually track all their client work in 6-minute increments. PointOne uses AI to completely automate time tracking and bill review, helping law firms collect more revenue and saving lawyers hours each week.

Lucite is building the first AI-native platform to help insurance brokers and advisors automate the most complex and time consuming parts of their job. We ingest unstructured carrier documents to accurately extract and analyze data, create insights, and generate deliverables, saving firms hours of time each day.

Rove is building the first accessible and universal airline mile, redeemable across dozens of international airlines and hotel chains worldwide.

Greenboard is building a singular, AI-powered platform to handle management of the entirety of financial firms’ compliance programs, and eventually all back office processes. Today, financial institutions use Greenboard to detect risk using AI and to rip out and replace the bad legacy back office tooling they have wanted to get rid of for years. Greenboard's mission is to enable financial institutions to operate their back office teams at a higher standard, with half of the headcount. We’re starting with compliance

Sendblue provides businesses with virtual iMessage lines to text prospects and nurture leads. Our users on average see a 70% increase in response rate. The company is founded by Rob Haber, former ML Engineer at Scale AI, and Alex Luis, former PM at PwC's telecom practice.

Fortuna is a one-stop solution that helps consumers navigate Medicaid end-to-end from eligibility and enrollment to renewing coverage. Our platform creates value as the middleware layer for the $800B Medicaid ecosystem.

Your Sales and Sales Ops team in a unified platform. No setup, Salesforce SSO and go.

Airgoods is building the future of food and beverage distribution. Today, we help emerging brands (think Fishwife, Ghia and David Protein) rapidly grow their wholesale business and earn placement into the nation’s best specialty and natural food stores.

Preventive heart health. Don't die of a heart attack -- use Empirical Health to measure 100+ biomarkers, predict your future risk of heart disease, and create a customized prevention plan with a doctor. We AI to deliver preventive health at scale. We're licensed, registered, and insured to deliver real medical care in 30+ US states covering 200m+ people.

ParadeDB is an open source, ACID-compliant search and analytics database. Built on Postgres for update-heavy workloads.

Cedana (YC S23) brings hyperscaler and frontier-lab orchestration capabilities for AI workflows. Our core capability is live migration for CPUs and GPUs workloads. This increases cost savings up to 80%, accelerates time to first token 2-10x, and enables stateful reliability of training jobs even through catastrophic GPU failures. We've integrated our solution into K8s, and support Kueue and Slurm for training distributed jobs, and Kserve for serving inference.   OpenAI, Meta and Microsoft have flavors of these capabilities internally and we’re bringing them to everyone.  Our vision is to transform cloud compute into a real-time, arbitraged commodity.  https://www.cedana.ai

Simbie AI voice agents are EHR-integrated and answer SMB medical practices' phones and call their patients for them. They handle both admin (scheduling) and clinical (results) tasks. Simbie is powering the return to independent medicine—50% of care delivered today—as doctors and patients badly want. After years of healthcare consolidation that left patients worse off and physician burnout surging, we're giving SMB the (infinite) scale to compete with hospitals, without the overhead.

Sohar Health is an AI-driven front-end RCM solution that transforms insurance verification processes for healthcare providers, enabling faster patient conversions and reducing administrative workloads. With a 95% automation rate, our API-first platform seamlessly integrates into existing workflows to provide real-time claim accuracy and eligibility checks. Key performance metrics showcase the power of our technology: Median latency of just 6 seconds, ensuring real-time eligibility; over 90% of checks returned within 30 seconds for increased patient conversion; 96% accuracy in identifying and verifying benefits details; Industry-leading 99% accuracy for eligibility determination; >90% carve-out detection rate, mitigating surprise billing risks; a 60% success rate with our Insurance Discovery API, helping identify coverage for self-pay patients. Our customers include outpatient clinics, specialty practices, and digital health platforms looking to streamline front-end claim management. By reducing errors and maximizing clean claim submissions, Sohar Health empowers healthcare organizations to convert and retain more patients, reduce operating costs, and increase their top-line revenue.