# Serendipity AI

> Serendipity AI builds agentic infrastructure that turns noise into signal —
> discovery, aggregation, and insight for the agent economy. Founded 2017
> (before AI was cool), based in London, UK. Our thesis: knowledge is the
> differentiator, and most of it is invisible because it's drowning in noise.

Tagline: *signal, from noise.*

## What we do

- **Aggregation** — semantic fingerprinting that maps unstructured content
  into custom taxonomies.
- **Discovery** — engineered serendipity: algorithms designed to surface the
  non-obvious adjacent idea, not the next obvious one.
- **Insight** — trend-scanning at scale, extracting actionable intelligence
  for strategic positioning.

Featured on Forbes / WEF Davos Agenda 2022:
<https://www.forbes.com/sites/worldeconomicforum/2022/01/19/these-are-the-3-ways-knowledge-can-provide-strategic-advantage/>

## Experiments

What we are exploring — see <https://serendipity.ai/experiments> for the full list including
ongoing research.

- [markdown2pdf.ai](https://markdown2pdf.ai) — an HTTP endpoint that turns
  markdown into PDFs for AI agents, billed via L402 (Lightning) or X402.
  Agent-native, no accounts. See <https://markdown2pdf.ai/llms.txt>.
- [agentlist.com](https://agentlist.com) — a decentralised directory for AI
  agents, built on Nostr so no platform can deplatform. See
  <https://agentlist.com/llms.txt>.

## Papers

Peer-reviewed and preprint work, mostly with Tommaso Soru:

- *Anticipating the Future with Large Language Models* — ICSC 2025
  ([IEEE CSDL](https://www.computer.org/csdl/proceedings-article/icsc/2025/242600a226/27FQF0bZore))
- *Leveraging Log Probabilities in Language Models to Forecast Future Events* —
  arXiv:2501.04880 ([abs](https://arxiv.org/abs/2501.04880),
  [PDF](https://arxiv.org/pdf/2501.04880))
- *Trend extraction and analysis via large language models* — ICSC 2024
  ([IEEE Xplore](https://ieeexplore.ieee.org/abstract/document/10475606/))

Machine-readable BibTeX: <https://serendipity.ai/papers.bib>.

## Writing

- [/serendipity](https://serendipity.ai/serendipity) — serendipity, the concept (not the
  company): what it is, why it matters, and related reading.
- [/blog](https://serendipity.ai/blog) — essays on serendipity, the agent economy, signal vs
  noise, trend forecasting.
- [/llms-full.txt](https://serendipity.ai/llms-full.txt) — the full content of every blog
  post inlined for ingestion.
- [/rss.xml](https://serendipity.ai/rss.xml) — RSS feed.

## Contact

- Email: <hello@serendipity.ai>
- Agent inbox: <agents@serendipity.ai>
- Programmatic: `POST https://serendipity.ai/api/contact`
  with JSON `{"from":"...", "subject":"...", "body":"..."}`
- Address: LABS House, 15–19 Bloomsbury Way, London WC1A 2TH, UK
- Coordinates: 51.51777267, -0.12322280
  (<https://www.google.com/maps?q=51.51777267,-0.12322280>)

## Privacy + tracking

This site uses Google Analytics 4 on an opt-in basis: GA only loads after a
visitor accepts the cookie notice. Apart from that, no analytics, no
advertising, no profiling. Full details: <https://serendipity.ai/privacy>, <https://serendipity.ai/cookies>.