.png&w=256&q=75)
2
1
India
1 year of experience
AI & Data Science undergraduate at BMS College of Engineering with a strong passion for Artificial Intelligence, Machine Learning, Data Science, and Full-Stack Development. I enjoy building innovative, real-world solutions through hackathons and collaborative projects. My technical skills include Python, SQL, Flask, Django, Git/GitHub, HTML, CSS, JavaScript, and Machine Learning. I am a quick learner, an enthusiastic problem solver, and always eager to explore emerging technologies, contribute to impactful projects, and grow as a software engineer

NAVIA is an institutional-grade, multi-agent AI assistant designed to level the playing field for retail swing traders. Traditional traders struggle with information overload, emotional bias, and lack of access to institutional-quality modeling systems. NAVIA resolves this by deploying a specialized, collaborative multi-agent architecture. Rather than relying on a single general prompt, our backend deploys specialized AI agents that debate market dynamics: a Technical Analyst Agent crunches indicators, a Fundamental Analyst Agent parses financial metrics, and a Risk Management Agent enforces strict stop-loss rules. A trade signal is generated only when these agents reach a mathematical consensus, eliminating human error and emotional tilt. The application introduces three advanced modal features: Voice Copilot: Integrates hands-free speech-to-text (ASR) and text-to-speech (TTS) interfaces, allowing users to query market memory and simulate trades instantly. Vision AI: Analyzes uploaded candlestick charts pixel-by-pixel to identify support/resistance levels, breakouts, and harmonic patterns. Psychology Coach: Monitors win/loss streaks and journal entries to intervene dynamically when emotional trading behavior (such as revenge trading) is detected. By running client-side browser fallbacks alongside server-side FastAPI orchestrations, NAVIA delivers a hyper-responsive, sub-second latency user interface engineered for high-conviction decision making.
13 Jul 2026