AlphaLog

The Vision

Most trading journals are spreadsheets in disguise — rigid data entry tools that ignore the psychological dimension of trading. AlphaLog was born from a simple question: what if your trading journal could actually think with you?

AlphaLog is an AI-driven trading journal and stock research platform that connects to 28+ brokerages, automatically imports trades, and applies artificial intelligence to detect behavioral biases, generate testable strategies from natural language, and provide research-grade analysis — all in one place.

As sole founder, I designed and built AlphaLog end-to-end: product strategy, UX/UI design, design system, full-stack engineering, and AI feature architecture. This case study tells the story of building an AI-native product — both how AI tools shaped the design and development process, and how AI features power the product itself.

Design Challenges

AI-Native Design Process

AlphaLog wasn't just built with AI features — it was built with AI tools at every stage of the design and development process. From ideation through production, Claude, Cursor, and other AI tools fundamentally changed how I work as a designer and developer.

How AI Shaped Each Phase

The AI Toolkit

A modern AI-native workflow combines specialized tools at each stage. Here's the stack that powered AlphaLog's design and development.

Portfolio Dashboard

The dashboard aggregates multi-account portfolio data into a unified net worth view with interactive charts, position breakdowns, and real-time performance tracking. A key design challenge was presenting data from 28+ different brokerage APIs in a consistent, scannable format.

The solution uses a card-based layout with progressive disclosure — summary metrics at the top, with drill-down into individual accounts and positions.

Trade Journal

The trade journal is the core of AlphaLog. Unlike spreadsheet-based journals, it automatically imports trades from connected brokerages and supports complex multi-leg options strategies — verticals, iron condors, straddles — without manual data entry.

Each trade entry captures quantitative data (entry, exit, P&L) alongside qualitative context: the trader's thesis, emotional state, and market conditions. This combination is what powers the AI features downstream.

AI Trade Audit

The AI Trade Audit is AlphaLog's signature feature. After a trade is closed, the AI analyzes the full context — entry timing, position sizing, market conditions, the trader's stated thesis, and journal notes — to detect behavioral biases.

The audit identifies patterns like FOMO, revenge trading, overconfidence bias, anchoring, and loss aversion. Rather than simply flagging issues, it explains the reasoning and suggests specific behavioral adjustments.

Design Principles for AI Feedback

• Explanatory, not judgmental — every bias detection includes the AI's reasoning chain
• Actionable suggestions paired with each finding
• Confidence indicators so traders understand the AI's certainty level
• Pattern tracking over time to show behavioral improvement

Strategy Lab

Strategy Lab lets traders describe strategies in natural language — “Buy when RSI drops below 30 and MACD crosses above signal line” — and converts them into backtestable logic. The AI handles the translation from human intent to programmatic strategy.

Backtesting results are presented with clear visualizations: equity curves, drawdown charts, win rate, and risk-adjusted returns. The design prioritizes making statistical results accessible to traders who aren't quants.

Trade Replays

Trade Replays let traders review closed trades bar by bar, re-experiencing the price action as it unfolded. Combined with journal notes and the AI audit, replays create a complete learning loop: what happened, what you were thinking, and what the AI observed about your behavior.

Research Suite

The Research Suite provides institutional-quality analysis tools for retail traders. The centerpiece is a 100-point Stock Intelligence scorecard that evaluates stocks across four pillars — Business Quality, Financial Strength, Earnings Quality, and Valuation — combining fundamental analysis with sentiment data and earnings tracking.

AlphaLog AI

The AI assistant is journal-aware — it has full context of the trader's history, positions, performance, and behavioral patterns. Traders can ask questions like “What's my win rate on earnings plays this quarter?” or “Which of my biases has improved most?” and get answers grounded in their own data.

The assistant also integrates through MCP (Model Context Protocol), allowing it to connect with external tools like Claude Desktop and Cursor while maintaining the journal context that makes its responses uniquely valuable.

Technical Architecture

AlphaLog is built as a monorepo with a modern AI-native stack. The architecture was designed for a solo founder to move fast without sacrificing production quality.

Design System

AlphaLog's design system uses a dark-first approach optimized for extended screen time — traders often monitor positions for hours. The color palette uses high-contrast accents for actionable elements, muted tones for dense data, and semantic colors for P&L (green/red) that remain accessible in both light and dark modes.

The component library is built with Tailwind CSS's design token system, enabling rapid iteration while maintaining visual consistency. Components were designed for data density — financial interfaces require more information per screen than typical consumer apps.

Key Solutions to Design Challenges

Lessons Learned

Building AlphaLog as a solo founder with AI tools has fundamentally changed how I think about the design-development boundary. AI doesn't replace design thinking — it amplifies it. The speed gains from AI pair programming and research synthesis freed me to spend more time on the problems that matter most: understanding user psychology, designing information hierarchy, and crafting interaction patterns that reduce cognitive load.