Algo Trading With
AI-Powered Strategies
A full lifecycle trading system combining rule-based strategies, LLM-driven research, and real-time execution across broker integrations.
Overview
We supported the development of an end-to-end algorithmic trading platform designed to help transform investment ideas into executable strategies.
The system enables both rule-based strategies using technical indicators and statistical logic, as well as more advanced LLM-driven workflows that incorporate unstructured data such as financial filings, news signals, and sentiment analysis.
Our role involved working closely with the client to explore hypotheses, structure ideas, and translate them into testable systems while contributing to architecture, pipelines, and execution layers.
LLM-Driven Financial Research
Intelligent workflows that dynamically retrieve, analyze, and reason over financial filings and market signals.
We built internal tools to fetch and process SEC EDGAR filings, including structured XBRL (XML-based) financial data. These tools extract relevant financial signals and normalize them for downstream analysis.
Using function-calling workflows, the system determines what information to retrieve, triggers scraping pipelines, and performs structured reasoning on top of extracted data.
Example: If a user requests analysis of shareholder information, the system first identifies the relevant filing (such as DEF 14A), retrieves it from EDGAR, extracts key sections, and then uses LLM reasoning to summarize ownership patterns, major stakeholders, and potential signals that could influence investment decisions.
This approach extends beyond a single query, enabling flexible analysis across filings, combined with external data sources like market data APIs and news sentiment pipelines.
System Capabilities
Strategy Research
Translate investment ideas into structured logic using indicators, statistical models, and sentiment signals.
Backtesting Engine
Validate strategies on historical data with robust evaluation, trade logs, and performance insights.
Simulation Layer
Realistic execution modeling including order fills, latency, and P/L tracking before live deployment.
Execution Systems
Automated pipelines integrated with broker APIs for seamless paper and live trading environments.
AI Analysis
LLM workflows for SEC filings, sentiment extraction, and unstructured financial data reasoning.
Monitoring & Iteration
Continuous tracking, reporting, and refinement of strategies based on performance metrics.
Technical Foundation
Core
Python
System & ResearchData
Market & News APIs
AlphaVantage, yfinance, NewsAPIAI
LLMs
Reasoning & AnalysisBrokers
Trading APIs
Alpaca, IBKR, Tastytrade