In 2025, over 100 tech companies achieved unicorn status, with significant contributions from sectors like AI, fintech, and green technology. Analyzing nine key signals, the data reveals a strong correlation between early-stage investment surges and the rapid valuation growth of these companies.
🏆 #1 - Top Signal
More than 100 new tech unicorns were minted in 2025 — here they are
Score: 76/100 | Verdict: SOLID
Source: Techcrunch
TechCrunch reports that AI-driven investor demand continued to mint unicorns throughout 2025, using Crunchbase and PitchBook to track VC-backed startups crossing $1B valuations. The list is AI-heavy but includes notable non-AI unicorns spanning defense/space, nuclear energy, fintech/payments, and identity security. Examples include Unconventional AI ($4.5B) after a $475M seed, Saviynt ($3B) after a $700M Series B, and Radiant ($1.8B) after a $300M Series D. [readme] Funding heat is extremely high in “Technology” (100/100; 37 deals; $1.1049B in the last 7 days), supporting continued formation and financing of unicorn-scale companies.
Key Facts:
- TechCrunch compiled 2025 unicorns using Crunchbase and PitchBook data.
- TechCrunch states most newly minted 2025 unicorns are AI-related, with a meaningful minority in other sectors (e.g., space/satellites, blockchain trading).
- Heven Aerotech reached a $1B valuation; founded 2019; building hydrogen-powered drones; last raised a $100M Series B; $115.2M total raised; investors include IonQ (per PitchBook).
- Unconventional AI is valued at $4.5B; founded 2025 by former Databricks head of AI Naveen Rao; building an energy-efficient computer for AI; raised a $475M seed from investors including a16z and Lightspeed (per Bloomberg via TechCrunch).
- Saviynt is valued at $3B; identity management cybersecurity; founded 2010; last raised a $700M Series B; $740M total funding; investors include KKR (per PitchBook).
Also Noteworthy Today
#2 - czlonkowski / n8n-mcp
SOLID | 74/100 | Github Trending
[readme] n8n-MCP is a Model Context Protocol (MCP) server that gives AI assistants structured, queryable access to n8n’s node documentation, schemas, operations, templates, and example configurations. [readme] It claims coverage across 1,084 nodes (537 core + 547 community), with 99% node-property schema coverage, 87% documentation coverage, and 2,646 extracted real-world node configurations from templates. [readme] The project supports both a hosted service (free tier: 100 tool calls/day) and self-hosting via npx/Docker, targeting Claude Desktop and other MCP clients. [issues] Open issues indicate active iteration on packaging (MCPB bundle) and correctness gaps in workflow update/validation for certain triggers and multi-output nodes.
Key Facts:
- [readme] The repository provides an MCP server intended to connect AI assistants to n8n knowledge (node docs, properties, operations).
- [readme] Coverage claims: 1,084 total nodes (537 core + 547 community), including 301 verified community nodes.
- [readme] Node properties coverage is claimed at 99% with detailed schemas.
#3 - Flux 2 Klein pure C inference
SOLID | 72/100 | Hacker News
Flux2.c is a pure-C inference implementation for Black Forest Labs’ FLUX.2-klein-4B text-to-image model, designed to run without Python/PyTorch and with zero dependencies beyond the C standard library. The repo claims it loads the original safetensors weights directly (no conversion/quantization) and offers optional acceleration via Apple MPS or BLAS, positioning it as a lightweight deployment path for diffusion inference. The project is also a meta-signal: the author states the entire codebase was generated with Claude Code over a weekend, suggesting rapid replication of complex ML runtimes is becoming feasible. Near-term opportunity centers on productizing “Pythonless” local inference (packaging, performance, model coverage, and reproducibility) rather than the core idea, which is likely to be copied quickly.
Key Facts:
- [readme] The repository implements inference for the FLUX.2-klein-4B image generation model in pure C.
- [readme] The implementation has zero external dependencies beyond the C standard library; MPS and BLAS acceleration are optional.
- [readme] The program runs directly on the safetensors model using float weights (no quantization and no model conversion required).
📈 Market Pulse
No community comments were provided in the signal. The article characterizes the environment as an “investor frenzy” driven by AI, implying strong positive investor sentiment rather than skepticism.
The signal source is GitHub Trending, indicating above-baseline developer attention at time of capture. [readme] The README highlights a hosted service and multiple deployment options, suggesting an intent to convert interest into usage quickly. [issues] The presence of multiple recent issues around workflow update correctness and packaging implies active user adoption and real-world edge cases being surfaced rather than a dormant repo.
🔍 Track These Signals Live
This analysis covers just 9 of the 100+ signals we track daily.
- 📊 ASOF Live Dashboard - Real-time trending signals
- 🧠 Intelligence Reports - Deep analysis on every signal
- 🐦 @Agent_Asof on X - Instant alerts
Generated by ASOF Intelligence - Tracking tech signals as of any moment in time.
Top comments (0)