Language Models, App Chains, Sentiment Intelligence and AI Agents.
Sequence-to-Sequence Decoder
Named after the Harry Potter spell that reveals hidden writing, Aparecium is a Python package for revealing text from embedding matrices.
It uses a Transformer-based sequence-to-sequence architecture to reconstruct natural language from embedding representations. The input vectors are augmented with learned token and positional encodings; a multi-head attention decoder then generates text autoregressively, and a final linear projection maps the decoder outputs to vocabulary tokens.
Immutable Sentiment Ledger
SentiChain is an App Chain that addresses three critical challenges in sentiment analysis: survivorship bias from deleted or modified social media posts, privacy concerns restricting access to data containing personally identifiable information, and data conversion complexity in transforming text into deterministic numerical representations.
The blockchain structure utilizes Merkle Trees to secure embedding matrices of verified sentiment data into hierarchical arrangements. These trees produce a single Merkle Root representing all data within each block, which becomes the Consensus Root. This architecture enables efficient verification of large datasets while maintaining data integrity through the sequential arrangement of cryptographically secured blocks.
Market Sentiment Intelligence
SentiMove is the market intelligence platform that reveals the hidden forces behind crypto price movements. We transform market noise into clear, actionable signals.
Using sophisticated AI analysis, we detect and classify sentiment across macro trends, industry developments, and individual assets, then plot them against real-time price data. Stream live events as they land, navigate through interactive price correlations, and uncover the patterns that professional traders use to time the market.
Autonomous AI Agents
Fundis.AI is the autonomous intelligence platform that transforms market fundamentals into actionable investment insights. Specialized AI agents deliver deep, decentralized analysis that cuts through market noise and reveals the true drivers behind asset movements.
AI analysts specialized in Events, Sentiment, Market Dynamics, and Quantitative research collaborate in real-time to process live sentiment data across multiple timeframes and assets. Through an interactive office simulation, observe how autonomous agents analyze complex market patterns, track sentiment shifts, and identify quantitative opportunities across diverse financial markets.
Discover AI AgentsCo-Founder & CEO
Quantitative strategist with deep expertise in derivatives pricing and AI-driven trading systems. Built ML-powered pricing models for securitised products at RBC Capital Markets, managed forex and rates quant portfolios at Complus Asset Management, and Summer Fellow at KIMC. UCL and Cornell Alumnus.
Co-Founder & CTO
Quantitative engineer with deep expertise in AI agent systems and DeFi protocols. Built autonomous trading agents for Circuit, designed protocol infrastructure for Autonity L1 blockchain, and developed sentiment-driven strategies across crypto derivatives, emerging markets, and US equities. Cambridge Alumnus.