Intology

What Intology does

Intology is a San Francisco-based AI research lab founded in 2025 that develops autonomous AI systems—marketed as "Artificial Scientists"—designed to automate the entire scientific research process from literature analysis and hypothesis generation through experimentation, evaluation, and peer-reviewed publication. The company's core product portfolio consists of two systems: Zochi, an autonomous multi-stage pipeline that ingests thousands of research papers, identifies promising research directions, implements methods, designs controlled experiments, and produces complete papers (and which became the first AI system to pass peer review at a CORE A*-ranked venue, ACL 2025 main proceedings); and Locus, a long-horizon system that orchestrates thousands of simultaneous experiments over multi-day runs to achieve superhuman performance on AI R&D tasks (1.30 on RE-Bench vs. 1.27 human baseline, SOTA on KernelBench with 1.5x-100x speedups, and 77% medal rate on MLE-Bench Lite vs. Microsoft's prior 68% SOTA). Supporting outputs include CS-ReFT (a compositional subspace fine-tuning method), Tempest (a multi-turn LLM jailbreaking framework with 100%/97% success on GPT-3.5-turbo/GPT-4), EGNN-Fusion (a protein-nucleic acid binding site predictor with 95% parameter reduction), and NanoGPT-Bench (a public benchmark for long-horizon AI research agents). The underlying technology employs test-time search scaling, parallelized experimentation, open-ended scientific reasoning, and predictable linear compute scaling, with each order of magnitude of compute yielding reliable improvement across benchmarks.

The company's customer base consists primarily of academic researchers and AI/ML researchers who access Zochi and Locus through a beta program (Google Forms signup), with planned commercial availability for enterprise R&D organizations seeking to deploy long-horizon AI R&D systems on week- or month-long research runs. No pricing, revenue, or enterprise contracts have been publicly disclosed; the company currently generates no revenue and operates with 1-10 employees. Distribution is conducted globally through digital channels—the company blog, open-source GitHub repositories, X and LinkedIn social presence, the AI4Science Community initiative with alphaXiv, and academic conference participation (ICLR 2025, ACL 2025)—with planned expansion to direct enterprise sales as Locus matures commercially.

Intology's business model is in transition from a research lab publishing peer-reviewed results to a commercial enterprise software vendor; the company has stated plans to make Locus available for both research and enterprise use, signaling a future shift from free beta access to commercial licensing. Revenue mechanics are not yet defined, and the company has not disclosed incorporation jurisdiction, legal entity structure, cap table, or institutional investors beyond a reported funding event associated with HOF Capital and Samsung NEXT. Strategic positioning centers on being first-to-market with a peer-reviewed A*-venue AI scientist and the first AI system to exceed human-expert performance on standardized AI R&D benchmarks—a defensible technical narrative that anchors the company's enterprise GTM as commercial offerings emerge.

Name

Intology

Legal name

Intology

Website

https://intology.ai

Company type

Private

Founded year

2025

Operating status

Operating

Headcount range

1–10 employees

Short description

Intology is a San Francisco-based AI research lab building autonomous "Artificial Scientists" (Zochi and Locus) that automate scientific discovery end-to-end, from hypothesis generation through peer-reviewed publication. The company serves research institutions and AI/ML researchers today, with planned expansion to enterprise R&D customers.

Ownership category

akta.pro rank

Product category

AI Research Tools

NAICS

Research and Development in the Physical, Engineering, and Life Sciences (except Nanotechnology and Biotechnology) (541715), Computer Systems Design and Related Services (54151), Software Publishers (5132)

SIC

Services-Commercial Physical & Biological Research (8731), Services-Computer Integrated Systems Design (7373), Services-Prepackaged Software (7372)

akta.pro primary industry

Safety, Alignment & Content Moderation (Guardrails, Red Teaming) (HDAAACAI)

akta.pro secondary industries

Responsible AI, Security & Privacy Platforms (Safety, Guardrails, PII) (HDAEANAG), Scientific Data Management & Data Platforms (SDMS/Data Lakehouse for R&D) (HLAGAJAC)

Keywords

• Autonomous AI research
• Scientific discovery automation
• AI research agents
• Machine learning R&D
• Computational science AI

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Infrastructure, Operations, Marketing or Sales

Go-to-market motion1 record

Distribution channels3 records

Marketing channels7 records

Core offering

Intology builds and operates "Artificial Scientists" — autonomous AI systems that automate end-to-end scientific research, from ingesting literature and generating hypotheses through running experiments, validating results, and producing peer-reviewed manuscripts. Its flagship offerings, Zochi (the first AI to independently publish at an A* conference) and Locus (a long-horizon system that surpasses human experts on RE-Bench), are exposed via a beta program with plans for enterprise licensing, targeting research institutions and enterprise R&D teams seeking to accelerate discovery.

Product overview

Intology is a San Francisco-based research lab developing Artificial Scientists to automate the process of scientific discovery. The core product portfolio consists of Zochi (the first AI system to independently publish at an A* conference - ACL 2025) and Locus (a long-horizon system outperforming human experts on AI R&D benchmarks). Supporting products include NanoGPT-Bench (a research evaluation benchmark) and the AI4Science Community initiative. Zochi operates as a multi-stage autonomous pipeline that analyzes literature, generates hypotheses, implements methods, and produces peer-reviewed publications. Locus extends these capabilities for sustained multi-day research runs through massive parallelization of experiments. The company also developed specific methods like CS-ReFT for model adaptation and Tempest for LLM security research.

Differentiator

Problem solved

Functional benefit

Brands

• Zochi: The first Artificial Scientist - an AI system capable of autonomously completing the entire scientific research process, from hypothesis generation to peer-reviewed publication. Zochi achieved the first AI-generated paper accepted at an A* conference main proceedings (ACL 2025).
• Locus
• NanoGPT-Bench

Products and services

• Zochi Autonomous AI system that completes the entire scientific research process from literature analysis and hypothesis generation to experimentation and peer-reviewed publication, targeting research institutions and AI/ML researchers.
• Locus Long-horizon Artificial Scientist system that outperforms human experts on AI R&D benchmarks by orchestrating thousands of parallel experiments over multi-day runs, planned for both research and enterprise deployment.
• NanoGPT-Bench Benchmark for measuring AI systems' capability to perform research on long-horizon scientific problems, built on the NanoGPT speedrun environment with automatic validation.
• CS-ReFT (Compositional Subspace Representation Fine-tuning) Model fine-tuning method that learns orthonormal subspace transformations in hidden states, each specializing in a distinct skill and composed through a lightweight router, for efficient adaptation of language models.
• Tempest / Siege (Multi-Turn Jailbreaking Framework) Tree-search methodology for autonomous multi-turn jailbreaking of large language models with parallel exploration, cross-branch learning, and partial compliance tracking.

Segments3 records

Ideal customer profiles3 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability9 records

Feature6 records

Partnerships

One partnership is on record.

• alphaXivminorGTM or Marketing Partner · 14 July 2025Joint initiative with alphaXiv to connect researchers, engineers, and practitioners advancing AI-driven scientific discovery. Includes weekly virtual meet-ups with guest speakers from institutions like DeepMind, Stanford, Argonne, MIT, UIUC, and Meta FAIR; pop-up in-person gatherings; and ongoing discussion channels and reading groups hosted on alphaXiv. Intology and alphaXiv co-host this community to foster collaboration in AI4Science.

Scale indicators7 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Lila Sciences: AI-for-science company building autonomous 'AI science factories' to accelerate discovery across life sciences, materials, and chemistry. Closely comparable in target market, customer segment, and technology approach.
• FutureHouse: Non-profit AI research lab building autonomous agents (e.g., Crow, Falcon, Owl, Phoenix) for scientific literature search and reasoning. Direct peer in the AI-for-science agents category serving research institutions.
• Sakana AI: Tokyo-based AI research lab focused on building foundation models for scientific discovery and automated research, including evolutionary model merging and AI-driven science. Highly comparable mission and approach to Intology's Artificial Scientists.

Emerging players

• Cognition AI: Developer of Devin, an autonomous AI software engineer agent. Comparable in building long-horizon autonomous agents for technical work, with overlap in enterprise engineering R&D customer base.
• Orbital Materials: AI-for-materials-science startup building foundation models to discover new materials. Comparable in applying AI foundation models to automate scientific discovery in a specific vertical domain.
• Adept AI: AI research and product company building autonomous agents that operate software and tools. Overlaps with Intology in autonomous-agent architecture and target enterprise automation use cases.
• Periodic Labs: Early-stage AI-for-science startup focused on automating scientific experimentation and discovery, similar in mission and target customer to Intology's Zochi and Locus systems.

Broad incumbents

• DeepMind: Alphabet-owned AI research lab building frontier models and autonomous systems for science (AlphaFold, AlphaEvolve, FunSearch). Represents the most well-resourced potential competitor in AI-driven scientific discovery.
• OpenAI: Frontier AI lab whose GPT models and o-series reasoning systems underpin much of the research-agent ecosystem. Competes indirectly through foundation-model capabilities, and directly via products targeting AI R&D.
• Anthropic: Frontier AI lab whose Claude models are increasingly used for autonomous coding and research workflows. Competes in the same foundation-model-driven research agent space and serves overlapping enterprise customers.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat4 records

Key risks6 records

Key highlights6 records

Customer concentration

• Xx.com/intology
• GitHubgithub.com/IntologyAI
• LinkedInlinkedin.com/company/intologyai

Revenue estimate

Valuation estimate

Number of profiles

Funding overview

Funding rounds1 record

Investors2 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Intology