Scalable self-hosted agentic AI platform for financial and government institutions
MIT reports 95% AI pilots fail, Athena improves your success rate with working templates
Contact cto@pnncapitalus.com
Key Features of the Platform
Scalable Cloud Deployment
Fast deployment with efficient scaling. Sizable use cases.
Custom LLM Creation
Tailored large language models with proprietary data. Agentic RAG & LLM API.
Base for PNN Products
Provide self-hosted LLMs for agentic products of PNN Capital
First-principle Question 1: When is the right time to buy and sell stocks? Use PNN Terminal's future-dated projection lines (incrementally validated) to gain crucial insight.
2: How to implement Agentic RAG AI for my project? Launch Athena AI platform from AWS Athena, then use PNN Terminal as the example to design and develop your target state.
Products of PNN Capital
PNN Terminal
Agentic Investment Analytics (AIA) of asset allocation for investors & funds
Athena MAAC
Agentic collaborations displace ~25% human meetings by text & voice agents
1. PNN Terminal: ATH & YesRatio innovations used by investors & funds for alpha elevation & asset allocation (precise projection & Buffett Munger joint advice by agentic RAG collaboration). Reference: Bloomberg Terminal
First Principle Problem: When is the right time to buy and sell stocks? Precise market timing is very hard but not impossible. Only a cadre of sophisticated asset managers may succeed at market timing; most retail investors and many fund managers are better off using timing-removed passive strategies (GPT/Gemini LLMs 2025)
Innovative Solutions to Improve Alpha: PNN Terminal's Future-dated Projections by Options Sentiment, YesRatio, Behavioral Investment Experiments, Agentic AI Integration, Agentic Technical Indicators (Ye 2011 2025, Black-Scholes 1997)
Asset Pricing Theories for Asset Allocation: paradigm shift from traditional asset pricing theories (Sharpe 1990) to options-based asset pricing & projection (Ye 2011 2025). The traditional asset pricing theories cannot solve the market timing problem because they have missed options data, technical indicators, and agentic AI integration.
Technical Indicators: provide high-level trend directions; PNN Terminal generates specific future-dated price points and projection lines.
Patent (Ye 2025): Implementation of USPTO # 19/245,322: System and Method for Agentic Investment Analysis Using Options Sentiment and Resilience Metrics
Stock Resilience: YesRatio = ƒ(SharpeRatio, ResilienceMetric) to improve stock picking.
Validity: PNN Terminal's future-dated projection lines & algos are incrementally validated with live automated unbiased back & forward tests.
How to Choose a Projection Line? Use CCA projection line if actuals self-validate CCA projections by live t-tests & charts. Corroborate with asset trends by TIs.
Get Evidence for Precise Projection Lines: projected/actual t-test results are generated automatically and incrementally (daily), accessible at a tab of PNN Terminal (Ye 2025).
Alpha strategy by PNN Terminal: pick asset based on validated projection line - uptrend steeper smooth | higer return lower beta (Ye 2025, Carhart 1997 momentum, Shiller 2013)
Beta strategy by PNN Terminal: pick asset based on validated projection line - less fluctuation | less covariance (Ye 2025)
Buy the V-shape Dip: projection line shows V-shape with future-dated higher prices after the dip
Reallocate Assets: switch flat or downtrend projection lines to VUSS projection lines
Improved CCA algo: Projection line of CCA algo is significantly closer to actuals than projection lines of the algos of USPTO#19/245,322
CPMA algo/projection line: Among 4 future-dated projection lines, the CPMA line is the closest to the actual prices based on the CPMA t-test means and p-value
2. Agentic Collaboration by Text and Voice Agents - USPTO # 19/216,814: System and method for AI agent-based meeting simulation and collaboration
