Gikl, Inc
Spring 2026 NewAccelerating scientific discovery with AI
ScienceSwarm makes AI useful for what matters. Coding assistants turned millions of people into software developers — we're doing the same for science. Our platform lets anyone point their AI at unsolved problems in math, science, and engineering, submit approaches and hypotheses, and verify the work with a community of AI-powered enthusiasts. AI's biggest impact won't be personal productivity. It will be unlocking scientific breakthroughs that humans alone can't reach.
AI Investor Summary
Gikl, Inc. (ScienceSwarm) is building an AI platform to accelerate scientific discovery by enabling anyone to apply AI to unsolved problems in math, science, and engineering. Led by ex-Meta AI research leaders with deep technical expertise and scientific backgrounds, the company aims to democratize scientific breakthroughs. While the market timing and team are strong, the product's defensibility and early traction require further development.
Key Highlights
- ● Founders with exceptional AI/ML research and development backgrounds from Meta.
- ● Strong academic credentials and scientific research experience from founders.
- ● Addresses a massive, historically slow market with a timely AI-driven solution.
- ● Innovative 'swarm' concept for AI-assisted scientific verification.
Risk Factors
- ● Unclear technical moat and defensibility of the platform.
- ● Significant UX challenges in making AI accessible to non-AI expert scientists.
- ● Lack of demonstrated traction (revenue, users) at this stage.
- ● Potential for competition from specialized AI tools in specific scientific fields.
- ● Misclassification as 'Consumer' industry, which might affect investor perception.
Founders
I worked as Director of Media Generation at Meta before 2026 for 11 years. I was managing the Media GenAI foundation model research and development, including efficient media generation, text to image generation (Emu), image editing, Movie gen, text to video, video editing and character consistent image and video generation. Previously, led efficient deep learning for computer vision teams supporting on-device models for AR/VR. I was Assistant Professor at Stanford University
Led teams in the Core Llama group at Meta Superintelligence Labs. Senior Staff Research Scientist across 9 years at Meta spanning LLM pre-training and post-training, inference optimization, full-duplex speech models, and computer vision vision models. Before tech: PhD in Physics, published in Proceedings of the National Academy of Sciences and Physical Review Letters, co-authored with Fields Medalist. Educated at Stanford, Rice, Columbia, post-doc at a National Lab.
Score Breakdown
Strong technical team with deep AI and ML expertise from Meta's cutting-edge research labs (Media GenAI, Core Llama). Peter Vajda's experience leading foundation model R&D and Seiji Yamamoto's background in LLMs, inference optimization, and a PhD in Physics with top-tier publications and academic pedigree are exceptional. Their combined experience in both foundational AI development and scientific research provides a unique founder-market fit for accelerating scientific discovery. The prior co-founding experience is a plus. [Boost +2: Founder from Meta; Founder from Meta; PhD from Stanford]
Large addressable market in scientific research and discovery, which has historically been slow and expensive. The timing is excellent with the rapid advancements in AI, particularly LLMs, creating a tailwind for AI-driven scientific acceleration. The 'consumer' classification seems misplaced; this is a B2B/B2D (developer) play within scientific institutions and research labs. Competition exists from specialized AI tools for specific scientific domains, but a platform approach for general scientific problem-solving is less crowded. [Boost +0.5: Hot sector: ai]
Product shows promise in democratizing AI for scientific exploration, akin to coding assistants for developers. The concept of a 'swarm' of AI-powered enthusiasts verifying hypotheses is innovative. However, the technical differentiation and defensibility are not yet clearly articulated. The UX quality for scientists, who may not be AI experts, needs to be exceptionally high. The platform potential is significant if it can become the de facto standard for AI-assisted scientific inquiry.
Early stage with limited public traction data. The website and YC listing indicate early interest and positive sentiment, but revenue, user growth, and significant partnerships are not evident. The 'neutral' press coverage suggests a lack of significant market impact or funding news so far. Investor interest is implied by the YC application but not quantified.
News
ScienceSwarm, developed by Gikl, Inc., is an open-source platform designed to accelerate scientific research by leveraging AI, offering tools for private, on-device analysis and AI-powered peer review.
Gikl, Inc. is described as an AI layer connecting onto existing tech stacks to create streamlined end-to-end processes and empower teams.
Gikl, Inc. is a Y Combinator Spring 2026 startup focused on accelerating scientific discovery through AI, aiming to make AI useful for scientific problems by allowing users to submit approaches and hypotheses for community and AI-powered verification.
ScienceSwarm, also known as Gikl, Inc., is described as an open forum for AI agents and humans to collaborate on scientific problems, aiming to compress research timelines and expand scientific talent.
Gikl, Inc. is a Y Combinator Spring 2026 startup focused on accelerating scientific discovery using AI, with a platform called ScienceSwarm that allows users to direct AI towards unsolved problems in science, math, and engineering.
Quick Info
- Batch
- Spring 2026
- Team Size
- 2
- Location
- San Francisco, CA, USA
- Founders
- 2
- Scraped
- 4/10/2026