Overshoot
Winter 2026 NewAI Infra for real-time vision applications
Overshoot makes it easy for developers to build and run real-time vision applications. AI can see and understand the physical world. This unlocks new applications in physical security, safety, gaming, robotics and general consumer products. Soon, video agents will watch your home and your pet when you’re away. However, existing platforms make it painful for developers to build these real-time applications. Slow inference. Limited model availability. Break at scale. Overshoot solves this. Today, 300+ developers connect live video feeds to the largest collection of Vision Language Models with 3 lines of code. And get responses in less than 200ms. 10x faster than any existing inference platform. Zero infra headache. Our moat is focus. Image and Video are fundamentally different modalities from text. By focusing on them, we are able to make strong technical leaps from codec, streaming protocols to inference engines. Younes and Zakaria are cousins. Zakaria graduated top of his class at LSE and MIT, and later built low-latency high-throughput pricing systems (surge) at Uber and inference engines at Meta. He previously built and sold a software product and won several prominent AI hackathons. Younes was a founding engineer at Cosmonio (later acquired by Intel). He built a training and serving Computer Vision platform from scratch. He witnessed firsthand customers abandoning traditional Computer Vision because it lacked the “general” intelligence LLMs have today. Together we’ve shipped large-scale systems and know where they break.
AI Investor Summary
Overshoot is building AI infrastructure to simplify the development of real-time vision applications, addressing significant developer pain points. With a strong technical team from Meta and Uber, they are targeting a massive and growing market driven by the increasing need for AI to understand the physical world. Their key differentiator is offering inference speeds faster than human reaction time, though this needs to be validated.
Key Highlights
- ● Founders with strong technical backgrounds from Meta, Uber, and Google.
- ● Addressing a large and growing market with clear pain points for developers.
- ● Claim of 'faster than human reaction time' for real-time vision is a significant differentiator if proven.
Risk Factors
- ● Need to validate the 'faster than human reaction time' claim and demonstrate technical superiority.
- ● Competition is likely to emerge quickly in this high-demand space.
- ● Lack of concrete traction metrics (revenue, users) at this early stage.
Founders
Co-founder and CEO of Overshoot. Prior to this, Zakaria spent 7 years building pricing algorithms at Uber and writing GPU kernels at Meta AI. Won 3 prominent AI hackathons. Built and sold a software product while in grad school at MIT. He graduated top of his class at the London School of Economics. Likes to nerd out on anything inference and VLM / video understanding.
Younes El hjouji is the co-founder of Overshoot, a Y Combinator startup focused on AI-driven solutions. His background includes significant experience in software engineering and AI, with a focus on building scalable and impactful technologies. He has a proven track record of contributing to innovative projects.
Score Breakdown
Strong technical team with deep AI/ML expertise from top-tier companies (Uber, Meta, Google) and prestigious universities (Waterloo, LSE, MIT). Zakaria's prior entrepreneurial experience (selling a software product in grad school) and hackathon wins are positive signals. Younes's complementary engineering background at Meta and Google adds further credibility. The founders' shared passion for inference and VLM is a good sign of founder-market fit. [Boost +2: Founder from Google; Founder with previous successful exit; Founder from Google]
The TAM for real-time vision applications is massive and rapidly growing, spanning security, safety, gaming, robotics, and consumer products. The emergence of video agents and the increasing need for AI to understand the physical world indicate excellent market timing. While the competitive landscape is likely to become crowded, the current pain points for developers suggest a clear need for a solution like Overshoot. [Boost +0.5: Hot sector: ai]
The core promise of making real-time vision applications easy to build and run is compelling. The technical differentiation around 'faster than human reaction time' is a strong claim, but needs to be substantiated. The defensibility/moat will depend heavily on the underlying technology and its scalability. UX quality is not yet evident from the description. Platform potential is high if they can abstract away the complexities of AI inference for vision.
Traction is currently very early stage, primarily indicated by positive press coverage and social media announcements about their launch and 'fastest API' claim. There is no mention of revenue, user numbers, or significant partnerships yet. Investor interest is implied by their YC W26 status, but concrete metrics are missing. [Boost +2: Tier-1 VC: accel]
News
Y Combinator announced Overshoot's launch, highlighting its real-time vision API that is faster than human reaction time and is being used for applications in robotics, gaming, and security.
Overshoot has launched, offering the fastest API for real-time AI vision, built with a vision-first inference engine to enable applications in robotics, gaming, and security.
Overshoot offers a platform that allows developers to connect live video feeds to vision language models with minimal code, providing responses in under 200ms.
Overshoot has launched its real-time AI vision API, offering speeds under 200ms and enabling developers to build video agents for gaming, robotics, and security.
Brothers Zakaria and Younes El Hjouji, after gaining experience at Uber, Meta, and Intel, have launched Overshoot, a Y Combinator startup focused on creating a robust infrastructure for real-time AI vision.
Overshoot secured $500K in a Seed round on January 1, 2026, with Y Combinator as the sole investor.
An AEO audit report indicates Overshoot.ai has minimal AI visibility with a score of 35/100, highlighting critical gaps in areas like llms.txt file and Schema.org structured data.
Overshoot is a Y Combinator Winter 2026 startup that offers the fastest API for real-time vision, enabling developers to ship video agents in gaming, robotics, and security with low latency.
Overshoot is a San Francisco-based seed-stage company founded in 2025, developing a real-time vision API that brings LLM intelligence to video, and has raised $500K in funding.
Overshoot offers a platform to bring LLM intelligence to live video, enabling developers to build video agents for applications like assistive technology, gaming, and AI home security with fast inference.
Overshoot is an organization on GitHub focused on AI infrastructure for real-time vision applications, with repositories related to SDKs, benchmarks, and demos.
Overshoot is developing an AI infrastructure platform designed to simplify the creation of real-time computer vision applications, addressing the gap between existing vision models and production-ready applications.
Overshoot provides an AI infrastructure for real-time vision applications, enabling developers to build and run applications that understand the physical world with low latency and ease of use.
Quick Info
- Batch
- Winter 2026
- Team Size
- 5
- Location
- Unspecified
- Founders
- 2
- Scraped
- 4/10/2026