
Overview traca is a split-view web platform combining live market dashboards with a conversational AI analyst. It delivers instant market intelligence through plain-language explanations of price movements and technical patterns, behavioral coaching that detects emotional trading signals like FOMO and revenge trading, and social automation via AI personas that generate platform-optimized content for LinkedIn and X. ✨ Core Features (MVP) Real-time Market Insights: Direct integration with Deriv API across Forex, Crypto, and Stock markets provides live pricing and historical data. Conversational AI: A unified chat interface handles market analysis queries, delivers behavioral feedback, and creates social media content in one seamless flow. Split-View Dashboard: Side-by-side layout presents live price charts, sentiment indicators, and AI analyst chat for efficient decision-making. Behavioral Pattern Detection: Advanced algorithms identify win/loss streaks, risk escalation patterns, and impulsive behavior, delivering timely nudges and habit-building reinforcement. Social Content Drafting: Platform-aware content generation produces professional LinkedIn posts and concise X updates tailored to each network's format. 🏗️ Architecture traca employs a Modular Monolith architecture optimized for real-time data streaming and AI inference: Frontend: React (Vite), TailwindCSS, shadcn/ui, and Zustand for state management. Backend: FastAPI (Python) with Uvicorn and WebSocket support for real-time communication. AI Engine: Mistral API cloud LLM for natural language understanding, market analysis, and content generation. Data Source: Deriv API integration for live market pricing, historical trade data, and account activity. Persistence: SQLite database stores trade history, session-based chat memory, and content drafts. This architecture ensures low-latency market data delivery, responsive AI interactions, and scalable performance.
7 Feb 2026

Nexus is an advanced, AI-powered academic tutor designed to function as a personal study agent for students. Its primary mission is to facilitate deep learning by intelligently processing and synthesizing information from a diverse setof course-specific materials uploaded by the user. This can include anything from dense textbooks and visual lecture slides to unstructured, handwritten notes, allowing Nexus to build a comprehensive understanding of the entire curriculum. Unlike general-purpose AI assistants that draw from the open internet, Nexus operates within a strictly defined knowledge boundary. Its responses are based solely on the provided content. This core feature ensures that every explanation, summary, and answer is directly relevant to the student's specific course, reflecting the unique perspectives, terminology, and scope defined by their professor. This eliminates the risk of students receiving confusing or contradictory information from external sources. When a student asks a question, Nexus doesn't just provide an answer; it delivers cited, multimodal responses. The "cited" aspect is critical for academic integrity and effective study: Nexus points directly back to the source material (e.g., "As stated in your 'Intro to Psych' textbook, Chapter 4, page 82," or "This concept is illustrated in your 'Week 5 Lecture Slides,' slide 12"), allowing students to verify information and review the original context. The "multimodal" capability means Nexus can generate explanations that go beyond simple text, potentially creating summaries, comparison tables, concept maps, or flashcards to present the synthesized information in the most effective way for learning.
19 Nov 2025