Infrastructure Startups funded by Y Combinator (YC) 2025

Fivetran automates data movement out of, into and across cloud data platforms. We automate the most time-consuming parts of the ELT process from extracts to schema drift handling to transformations, so data engineers can focus on higher-impact projects with total pipeline peace of mind. With 99.9% uptime and self-healing pipelines, Fivetran enables hundreds of leading brands across the globe, including Autodesk, Conagra Brands, JetBlue, Lionsgate, Morgan Stanley, and Ziff Davis, to accelerate data-driven decisions and drive business growth. Fivetran is headquartered in Oakland, California, with offices around the world. 

RLVR changed the game for post-training. We're here to build and supply the next-generation of coding data and RL environments using model checking languages and formal verification. In a world where solid coding data is becoming increasingly sparse, Haladir provides the data and environments needed to build the next generation of coding models.

Mantis provides the foundational infrastructure layer that enables companies to build human-in-computer models. We unify fragmented data sources, simulate human anatomy and physiology, and validate against real-world outcomes. From motion capture and biometric sensors to medical imaging and training logs, our platform transforms disparate data into validated digital twins ready for production applications.

Didit is an AI-native, developer-first identity verification platform that makes KYC, AML, and fraud prevention instant, flexible, and transparent, with a seamless user experience and more than 70% lower costs.

Luel is the go-to market for fast, rights-cleared multimodal training data at scale. We work with frontier AI teams to provide high-quality bespoke data collections and off-the-shelf datasets.

Captain delivers the most accurate general-purpose retrieval engine for unstructured data. Connect file stores and effortlessly retrieve knowledge with much higher accuracy than RAG (Avg: 78% → 95% + citations).

Castari lets developers deploy AI agents in secure, autoscaling sandboxes with one command. We handle tool/MCP connectors, snapshots, and observability so teams go from prototype to production in hours instead of weeks. Developers focus on building agents, Castari provides the autoscaling sandbox infrastructure to ensure your agents and its tools are secure.

Nivara gives engineering leaders visibility into how humans and AI build together, helping teams spot bottlenecks, measure impact, and turn AI adoption into real outcomes. We're building the much needed copilot for Engineering Managers and empower every team to work intelligently with AI.

Hyperspell (YC F25) gives AI agents memory. It connects to tools like Slack, Gmail, Notion, and Drive so agents can recall, reason, and remember across company knowledge. Developers use Hyperspell’s API to turn disconnected agents into AI coworkers that understand their work. Conor and Manu founded Hyperspell after building their own workplace agent and seeing how limited AI is without memory. They have over 15 years of experience in machine learning and have scaled API-based products to $30M ARR.

AI researchers should be pushing the boundaries of what's possible with new architectures and training methods. Instead, they waste weeks configuring cloud infrastructure, debugging distributed systems, and optimizing their GPU code. We know because we lived it: While training our own models across thousands of GPUs earlier this year, we spent more time fighting our infrastructure than doing actual research. That why we're building two things. First, Elastic Cloud: a managed platform that automatically finds the cheapest GPUs across all providers, handles spot instance preemption, and cuts compute costs by up to 80%. Second, automatic kernel optimization that makes training code run faster by modeling hardware topology, often beating hand-tuned implementations. The problem is that getting high performance across different hardware is genuinely hard. NVIDIA's CUDA moat exists because writing fast kernels requires deep expertise. Most teams either accept vendor lock-in or hire expensive kernel engineers. Our goal is to break the CUDA moat. The compute bottleneck is the biggest constraint on AI progress. NVIDIA can't manufacture enough GPUs, and their monopoly keeps prices astronomical. Meanwhile, AMD, Google, and Amazon are shipping capable alternative hardware that nobody uses because the software is too hard. We're breaking that moat. If we succeed, anyone will be able to train state-of-the-art models without thinking past their PyTorch code.

Dome makes it easy to build on prediction markets like Polymarket and Kalshi. We provide the developer infrastructure to access, analyze, and build on prediction markets. With our unified API, developers can access live and historical data, including granular datasets not available natively, across multiple platforms at once. Dome makes it simple to trade, embed market data into products, and deploy strategies across multiple platforms through one interface.

Narrative is the AI video understanding infrastructure. We let language models understand thousands of hours of video just as easily as they understand text.

S2 is a serverless datastore for real-time, streaming data. The more we worked on data systems, the more we felt like there was a missing building block. The seamless experience of object storage simply did not exist for durable streams – so we set out to fix that! S2 reimagines streams as an unlimited and access-controlled web resource. Like if Kafka and S3 had a baby. - Building agents? Use granular streams to persist context in real time, make workflows auditable, and coordinate between agents. - Publishing fast-moving data? Streams support high fanout and can be accessed directly over the internet, so they are a great fit for broadcasting real-time feeds. - Sandboxing? Emit to per-instance streams, and read history or live events just as easily. - Local-first and multiplayer experiences? Propagate updates reliably and efficiently. Stop dealing with clusters, and start streaming.

Cursor for construction estimation & project management. Brickanta analyzes estimates and project details to help automate admin work, reduce errors and inaccuracy, and accelerate delivery with AI trained on the standardized building protocols.

We build infrastructure enabling AI agents to reliably perform complex tasks in production

The first advertising infrastructure designed specifically for AI apps. Seamlessly integrate contextual brands & products that enhance user experience while generating revenue to pay the inference bills.

Truth Systems provides a compliance agent to govern the use of generative AI. It translates operational policies into intelligent, real-time guardrails that monitor every AI interaction. This empowers companies to proactively block non-compliant actions and align access to AI tools with user competency. By eliminating the risks and operational blind spots associated with uncontrolled AI adoption, our platform enables enterprises to fully leverage the technology's potential without compromising security. We are establishing the new standard for AI governance, building a future where every professional can innovate safely and confidently.

Kernel is crazy fast, open source infra for AI agents to access the internet 🧠🕸️ - Built on unikernels for crazy fast cold starts and low resource usage - Sub-150ms cold starts on managed headful browsers - Built-in stealth mode with residential proxies and automatic CAPTCHA solving - Full observability: session recording, live view, and full logs - Persistent sessions that maintain cookies and auth across invocations - Native support for all browser and computer use frameworks - SOC 2 compliant and HIPAA-ready Trusted by Cash App, Rye, and 1000+ others. Backed by Accel & Y Combinator (S25).

DeepAware builds an AI-driven automation system for GPU-intensive data centers. Our reinforcement-learning scheduler, real-time market integration, and unified dashboard slice energy waste by up to 30%. Coming soon: autonomous “robot-hand” inspections and maintenance to enable 24/7 operations with minimal staff. More info: https://deepawareai.com/ Jobs/Internship: https://www.deepawareai.com/careers

mcp-use provides open-source dev tools and infrastructure for MCP to help dev teams quickly build and deploy custom AI agents with MCP servers. Our library has over 170,000 downloads and 7,000 stars, and is used by hundreds of engineers at big enterprises. We provide a hosted and self-hosted platform that manages auth, access control, and multi-user environments for secure MCP agents deployment at scale.

Pangolin is building a new kind of open-source proxy: cloud-orchestrated, identity-aware, tunneled, and highly available. It’s the simplest way to securely expose applications on any network—only the people you want get access to your apps, and we block everyone else.

You built the demo in a weekend. Now you need auth, a prod-grade database, real-time events, AI observability, cron jobs, email, monitoring… and you needed them yesterday. We spent a decade at our previous startup, building a platform while scaling to millions of users and customers like Meta & Netflix. Now we put that into an open-source framework, so others wouldn’t have to do the same from scratch. Modelence is a batteries-included framework that abstracts away boilerplate so you can just focus on building your product logic and nothing else - it’s like Next.js + Vercel + Supabase on a single platform, built for full-stack TypeScript. Shipping production apps is much harder than building a prototype that runs locally and 95% of all production apps share the same building blocks. We did a massive amount of work on our previous platforms that was not specific to our product, and we’ve created Modelence to abstract that away: - Authentication - Database setup - Hosting / CDN - AI integration (LLM observability, prompts) - Real-time events - Cron jobs - Email service - Monitoring - Analytics - Admin dashboards Andrej Karpathy recently posted about this same exact problem: https://x.com/karpathy/status/1905051558783418370 and explained it further in his YC AI Startup School talk: https://youtu.be/LCEmiRjPEtQ?feature=shared&t=1940

Luminal builds an ML framework and compiler that generates GPU code. Our stack 10x's model speed while simplifying deployment and cutting idle GPU costs Github: https://github.com/luminal-ai/luminal Discord: https://discord.gg/APjuwHAbGy

AgentMail is the email inbox API for AI agents. It gives agents their own email inboxes, like Gmail does for humans.

Janus automates AI evaluations by using high-fidelity simulation environments, catching failures in reasoning, compliance, tool usage, and performance. The resulting datasets benchmark products and feed post-training loops to continuously improve performance over time.

Aravolta provides a single pane of glass software for data centers to increase uptime, reduce costs, and improve performance. A single utility node onboards every asset within an hour and detects hardware failures, hotspots, and other issues. It also generates reports on important metrics like Power Usage Effectiveness (PUE), energy metering, resource utilization, and much more.

ParaQuery is a fully-managed GPU-accelerated Spark solution providing double the performance at half the cost compared to solutions like Databricks and BigQuery. It is fully Spark-compatible, cloud agnostic, and has seamless integrations everywhere, all without vendor lock-in. ParaQuery makes it trivial to run large SQL and Spark workloads efficiently and serverlessly, without any data migrations.

Morphik is open-source multimodal search for AI apps. We're used by customers ranging from space-tech teams searching across research papers to developers building brokerage agents.

PgDog is an application for sharding PostgreSQL. It understands SQL and can distribute queries automatically between databases. It's built for managed databases, like AWS RDS, and doesn't require any changes to application code or schema. In addition to sharding, PgDog is a load balancer and pooler, so it can act as a replacement for PgBouncer, RDS Proxy, and other Postgres scaling products. It brings the simplicity and performance of HTTP load balancing to the database. Built from the experience of sharding Postgres at Instacart during peak growth in 2020, PgDog is the answer to the old question: does Postgres scale? It does now.

Open-source context retrieval for AI agents across apps and databases. It connects to productivity tools, databases, or document stores and transforms their contents into searchable knowledge bases for agents.

RunRL improves LLMs and AI agents with reinforcement learning. Do you have a model that needs to get better at a certain task? Are you tired of messing around with prompts? Do you spend a lot of money on observability, and wish that all this data could let your model self-improve? Well, head on over to RunRL.com, and see what we have to offer! If you give us a model, a prompt, and a reward, we’ll make your model’s reward go up.

AI models are terrible in non-English languages because it's nearly impossible to find training data in other languages. So, we're building the world's largest and highest-quality multilingual data library.

assistant-ui helps frontend developers add AI chat to their apps. Our product is used by several dozen companies, like LangChain, Stack AI, Browser Use, and Athena Intelligence. We're building an open source Typescript/React library and a backend as a service, with first class integrations into popular agent frameworks like LangGraph and Vercel AI SDK.

Our CLI makes ML model training effortless - just describe your job, and we handle GPU orchestration and execution at half the cost of major cloud providers.

AfterQuery is a research lab (backed by Y Combinator) investigating the boundaries and capabilities of artificial intelligence through novel datasets and experimentation. We believe that pushing the frontiers of AI requires both rigorous scientific inquiry and exceptional human data.

Flowtel is building the next generation operating system for hospitality, powered by modern AI agents that can handle everything from booking to room service.

Butter is building an LLM proxy that records and deterministically replays tool call trajectories. Our goal is to get LLMs out of the hotpath for repetitive tasks, increasing speed, reducing variability, and eliminating token costs for the many cases that could have just been a script.

We are on a mission to build the world’s most accurate search engine over complex and unstructured documents. Most AI products - whether Q&A bots or AI agents - depend on retrieval systems to provide relevant context from knowledge bases. Yet, the vast majority of these systems rely on basic semantic or hybrid search methods, which still frequently fail. These mistakes lead to inaccurate responses and hallucinations by LLMs, frustrating developers and end users alike. That is why we’re building ZeroEntropy: to add intelligence to retrieval and empower developers to create AI products that are more reliable and accurate.

Coblocks is a thoughtfully-designed data platform that helps teams write queries and automate workflows faster. We understand the columns, tables, and relationships in your data and use them to help anyone on your team build pipelines with AI, SQL and Python. Think of us like Zapier plus Cursor for data engineering. Here’s how we’re different: • All-in-one: You can get started in 2 minutes – no setup or configuration required. We have one-click integrations, warehousing, transformation, and schedules all built in. • Seamless integrations: Plug in your Postgres database, Stripe transactions, Hubspot leads, or any other data source, without writing code to keep things in sync. • Thoughtful AI: We love Cursor and we love data – we combined the two to help you write accurate queries. We use existing metadata to help you create new datasets, connect sources, fix errors, or edit in place. • Collaborative: Easily share data and discover what others in your org have built as a starting place for your analysis. Wrap common blocks of logic with templates so your team never has to start from zero. • Resilient and Scalable: Our compute engine is lightning-fast for queries and builds. Git and branching are built-in for both code and data, so you can time-travel backwards when things break. You can start with GBs and grow to TBs.

Reliability on autopilot. Free you team to build what's next.

Ryvn provides the devtools and infrastructure to easily deploy any workload, on any cloud, and on-prem. Run apps in your own cloud and offer Bring Your Own Cloud (BYOC) to your customers in minutes, with built-in observability designed for sensitive industries.

Enterprises are paralyzed by the risks of deploying generative AI. Traditional methods like RAG and fine-tuning are brittle, expensive, and fail to prevent costly hallucinations or guarantee compliance in high-stakes environments. This lack of trust stalls innovation and blocks ROI. CTGT ensures strict, auditable adherence to your policies on the fly. We deliver defensible audit trails and remediate noncompliant model behavior, so outputs become more aligned with company policies over time, combating data and model drift. CTGT is backed by Google's AI fund (Gradient), General Catalyst, and luminaries like Paul Graham. We provide the control and confidence necessary to unlock the full potential of enterprise AI.

Revyl is a AI-native mobile testing platform that automatically catches and triages bugs before they reach production.

You share production/eval traces for your AI software and sync your code. Synth asynchronously runs experiments to develop, improve, and maintain your software and hands code/prompts/models off to your team. Designed for teams building vertical AI software/agents.

Archil transforms S3 buckets into a 30x faster, unlimited, local disk. Archil enables AI, analytics, and serverless applications to instantly access massive data sets without waiting for data transfer. Researchers use Archil for shareable, local storage of data set and model versions that never runs out of capacity.

Self-serve GPUs in VS Code with persistent storage, snapshots, and hot-swappable hardware, with prices 80% lower than AWS.

Miru makes it easy for robotics teams to manage, version, and deploy robot configurations. Developers use our well-documented APIs and UI components to automate configuration updates for both engineers and operators. In the long term, we're building software infrastructure to accelerate the automation of the physical world.

Opslane is a developer-first Kubernetes delivery platform. Connect your repo and get previews, safe rollouts, and instant rollbacks — no YAML, no ops team required.

We build software that helps hardware teams quickly and clearly evaluate how a component (engine, battery, valve, etc.) or process (assembly line, manufacturing plant) performs. We prevent mistakes, reduce costs, and accelerate time-to-market. Our platform unifies industrial control with sensor data storage and automated post-processing, delivering an end-to-end operations and observability stack. To put it comparatively, take LabVIEW, InfluxDB, and Grafana, wrap them into a single package, add a bit of magic, and you get Synnax.