# FIM (飞沐 / FIM Labs)

> FIM builds AI products that actually run: Zico (self-hosted contract lifecycle management), FIM One (model-agnostic enterprise agent platform), and a line of research tools. Each product has its own site and real users. Entities: FIM Labs Pte. Ltd. (Singapore) and Beijing FIM Network Technology Co., Ltd. (北京飞沐网络科技有限公司, Beijing).

## Flagship products
- [FIM One](https://one.fim.ai): Model-agnostic enterprise agent platform. Dynamic DAG planning, a RAG retrieval pipeline, and a visual workflow editor. Works with any OpenAI-compatible model and deploys on-prem.
- [Zico](https://zico.fim.ai/): Self-hosted contract lifecycle management. Native Feishu integration, AI contract-risk review, and the full drafting-to-signing-to-archive flow, with private deployment.

## Other products
- [Cito](https://cito.fim.ai): Semantic paper search. Search papers by meaning, not keyword overlap: SPECTER2 fused with BM25 retrieval, covering ~148M papers.
- [Tuto](https://tuto.fim.ai): Citation-integrity auditing. Checks two layers of a citation: whether the cited work exists, and whether it supports the claim. First run: all 209,985 references in ACL 2026.
- [Aha.](https://aha.fim.ai): Plain-language arXiv paper explainer. Paste an arXiv link and read the paper side by side with a plain-language explanation.
- [Hido](https://hido.fim.ai): Anonymous code hosting for double-blind review. Turn a GitHub repo into a read-only mirror that does not leak author identity. Reviewers browse, download, and clone with no login. One anonymization pipeline, prerendered to R2 for static serving.
- [Build GPT-2 from Scratch](https://learn-llm.fim.ai/): Hands-on course: write your own GPT-2. A Mandarin course that rebuilds GPT-2 end to end, from tokenization to training, one step at a time.
- [Linear Algebra → Attention](https://l2a.fim.ai/): Visual path from linear algebra to attention. Interactive visualizations that walk from linear algebra all the way to the Transformer attention mechanism.
- [Algorithm Visualization Lab](https://algo.fim.ai/): Drag to see the geometry behind algorithms. Interactive algorithm visualizations: drag a handle, watch a sort or a graph traversal rearrange, and read off the geometric idea underneath. A Chinese living notebook that grows one algorithm at a time.

## Deployments (who runs FIM systems)
- Newborn Town Inc.: Contract Lifecycle Management (powered by Zico)
- CAR Inc.: Contract Lifecycle Management (powered by Zico)
- Tsinghua University: AI Writing Platform (powered by FIM One)
- Peking Union Medical College Hospital: Gynecology Medical Knowledge Graph
- Peking Union Medical College Hospital: POP Surgical-Approach Recommendation System
- Beijing Ditan Hospital / WHO Collaborating Centre: Specialty Follow-up Platform
- China Cancer Foundation: Grassroots Tumor Center Training Platform
- Peking University: Quantitative-Paleontology Platform

## Research
- AI as Equalizer or Amplifier? Task Complexity as the Moderating Factor for Human Expertise in Hybrid Intelligence Systems (HHAI 2026 · IOS Press, 2026) — https://arxiv.org/abs/2512.10961
- Cognitive Workspace: Active Memory Management for LLMs (Preprint · arXiv, 2025) — https://arxiv.org/abs/2508.13171
- When Should the Agent Speak? A Survey of Intervention Timing for Always-On AI Assistants (Under review · TMLR · Assigned AE · Survey, 2026) — https://doi.org/10.5281/zenodo.21438396
- Registers, Not Plans: What Lives in a Language Model's Workspace That Isn't on Its Tongue (Under review · EMNLP 2026 Workshop · BlackboxNLP, 2026) — https://openreview.net/forum?id=MNyJaBJ5Mx
- The Preference Centroid: Consensus Density Governs Output Dispersion in Aligned LLMs (Under review · TMLR, 2026) — https://openreview.net/forum?id=6ukieTMBcG
- Fidelity Before Structure: Verbatim Chunks Beat Lossy Artifact Extraction in Long-Conversation LLM Memory (Under review · ARR 2026 August → EACL 2027, 2026) — https://arxiv.org/abs/2601.00821
- CGEP: Toward Detecting and Attributing GEO Poisoning in Chinese AI Search (Under review · EMNLP 2026 Workshop · NLP4PI, 2026)
- A Graph-Neural-Network Method for Data-Information Recommendation (Patent · Pending, 2026)

## Site
- English: https://fim.ai/en
- 中文: https://fim.ai/zh
- Deployments: https://fim.ai/en/deployments
- Case study (Newborn Town on Zico): https://fim.ai/en/deployments/newborn-town-zico
- Research: https://fim.ai/en/research
- Articles (paper walk-throughs & practice notes): https://fim.ai/zh/articles
- GitHub: https://github.com/fim-ai
- LinkedIn: https://www.linkedin.com/company/fim-labs
- Contact: hi@fim.ai