Fahrenheit
Research
We build intelligence at the edge of what is known.
Our research becomes product.
01 // PORTFOLIO
Our products dont just enter
markets they define them
Kashew AI
AI powered prospecting, pipeline, and relationship intelligence. Turn signals into pipeline. Automate outreach at scale. Smarter deals, faster closes.
WebsiteNinth
Venture intelligence for smarter fundraising and introductions. Data-driven dealflow, simplified. Connect with the right investors. Build networks that compound.
Website02 // DOMAINS_OF_INQUIRY
domain of
inquiry
Our research agenda spans the hardest problems in applied AI from deep data infrastructure to emergent model behaviour.
Deep Data Architecture
Building proprietary data pipelines that surface signal from noise at scale, semantically indexing billions of data points.
Small Language Models (SLM)
Distilling frontier-level reasoning into models under 7B parameters that outperform 70B baselines on domain-specific tasks.
Autonomous Reasoning Architectures
Multi-step inference engines that decompose novel problems, form hypotheses, and self-correct in real time without supervision.
Synthetic Data Generation
Proprietary pipelines producing training data indistinguishable from real-world distributions, closing the data scarcity gap.
Multi-Agent Emergent Behavior
Mapping the boundary between programmed behavior and spontaneous cooperation to engineer systems that exploit emergent dynamics.
Compressed Intelligence
10x size reduction with under 3% accuracy degradation — large-scale capabilities on constrained hardware at production scale.
03 // OPERATING PRINCIPLES
We do not license our breakthroughs.
We do not outsource our thinking.
Every model we train.
Every system we build.
Every product we ship.
Is ours.
04 // DEEP DATA
data is
the moat
Proprietary data infrastructure and annotation power models built for real problems.
05 // CONTACT
work with
the lab
We collaborate with researchers, enterprises, and builders who are serious about AI. If you are building something that matters, we want to hear from you.