Agentic Sciences, Inc. · San Francisco

Voice in.
Trading agents out.

Every strategy traceable to a conversation.

I describe hypotheses to the agents — by voice or text — sourced from conversations with practitioners (brokers, traders, asset managers), market observations, papers. The agents formalize each hypothesis into a backtested strategy. Deployed live on founder capital. No black box, no demo theater.

Use cases Founder Live work
How it works

Strategies, traced to source.

Hypothesis in, live capital out — automated end to end. Two examples below; each traces back to a specific input I gave the agents (voice or text), sourced from a practitioner conversation.

01 Input → 02 Research → 03 Signals → 04 Execution
CASE 01
Research published · Paper trading live

Volatility selling on equity-index options

Source — a conversation with a brokerage trader on variance-risk-premium intuition: when implied vol systematically overshoots realized vol, and how to harvest the gap without blowing up in tail events. I described the heuristic to the agents; they formalized it.

Agent output — two backtested strategies, each with full trade logs, Greek exposures, and per-day position files.

  • US equity vol selling. 4.6M options, 6,043 securities, 2016–2021. Sharpe 1.10, 73% win rate, 1,400 trades over 70 months.
  • CSI 1000 vol selling. 68K options, 1,022 contracts, 2022–2026. +71.9% total return, Sharpe 0.94, 305 trades, 63% win rate.
Research detail CSI 1000 paper · 91pp US equity paper · 51pp
CASE 02
Research in progress

Bond relative-value across the Treasury curve

Source — a conversation with a US asset manager on persistent dislocations across the Treasury curve: where carry, roll, and richness/cheapness signals diverge from textbook fair-value models. I described the view to the agents; they systematized it.

Agent output — the practitioner's view turned into a quantitative cross-curve relative-value framework. Backtested historical dislocations, documenting where the heuristic worked and where it broke.

Bond RV research site

Practitioners are anonymized. Their intuition shaped the hypothesis; the agents did the formalization. No endorsement implied.

Built solo

One founder. Production infrastructure.

Fourteen years of quantitative-economics training, deployed as the engine that powers every use case above.

Qihong Ruan
Qihong Ruan
阮启宏
Founder · SF Bay Area
Personal site LinkedIn GitHub 𝕏

Sun Yat-sen University → Xiamen MA (Quantitative Economics) → Cornell PhD in Economics, defended July 2025. AFA 2026 paper accepted on perpetual-futures cross-market predictability — the academic version of the methodology that drives the agents.

Solo founder. Trading on personal capital. Building the platform that powers it.

By the numbers
440
Voice memos
processed
238h
Audio transcribed
at $6.65 total
729
Autonomous
research reports
4
Production
machines
860
CPU cores
across cluster
126TB
Research data
across 272 WRDS
Credentials & stack
Cornell PhD · Economics · defended July 2025 · 14 years of quantitative-economics training, single domain.
AFA 2026 · paper accepted on perpetual-futures cross-market predictability.
4× NVIDIA Professional Certified at GTC 2026 · Data Science (96%), LLMs, Agentic AI, AI Infrastructure.
Voice pipeline · four-source ASR ensemble (Apple SpeechAnalyzer, Whisper, Gemini Flash, pyannote diarization), reconciled by Claude Opus.
Working now

Live dashboards & research.

Every link below resolves. No mockups. The same agent infrastructure produces all of it.

Contact

Get in touch.

Pick the right routing — every alias lands in the same inbox for now.

contact@agenticsciences.ai
investors@agenticsciences.aiVCs · capital partners
careers@agenticsciences.aihiring · open from Q3 2026
press@agenticsciences.aimedia · interviews
qihong@agenticsciences.aifounder · 1-on-1