---

$ whoami

class Tim:
    def __init__(self):
        self.education  = "BSc Data Science @ UCL (First Class)"
        self.role       = ["Technology Analyst @ BlackRock AWT", "Founder @ TempThreads"]
        self.team       = "Client Delivery & Execution, London"
        self.passion    = "Agentic AI systems that reason, not just retrieve"
        self.chess_elo  = 1903  # FIDE rated & coach
        self.languages  = ["English 🇬🇧", "French 🇫🇷"]  # both native
        self.citizenship = ["UK", "FR", "US"]

⚡ What I'm Building

I'm obsessed with agentic AI — systems that don't just call an LLM in a loop, but actually plan, route, retrieve, verify, and self-correct. From founding a startup with multi-agent LLM pipelines to building RAG systems that catch when the Federal Reserve changes its mind, I care about AI that can explain every decision it made and prove it didn't hallucinate.

🏗️ Featured Projects

🔮 Kronos — Autonomous Algorithmic Trading Engine Python · Transformer Forecasting · Alpaca API · FastAPI Multi-asset algorithmic trading system built on a transformer-based K-line foundation model, generating long signals from 5-minute candles with a 120-step forward horizon. Extended with a news/macro sentiment veto gate, ATR-based tail-risk stops, and a live paper-execution engine via the Alpaca API across equities (NVDA, TSLA, MSFT, AAPL, AMZN), metals (GLD, SLV), and crypto (BTC, ETH, SOL). Currently in active backtest validation — auditing for data leakage and correcting annualization methodology before trusting reported Sharpe. Transformers Time Series Forecasting Quantitative Trading Alpaca API FastAPI
🏦 Agentic Financial RAG LangGraph · Pinecone · GPT-4o · FastAPI · RAGAS Agentic RAG system that analyses FOMC documents to detect monetary policy shifts and contradictions. 5-node LangGraph state machine with self-verification, dual embeddings, and RAGAS evaluation. Agentic AI LangGraph Pinecone FastAPI RAGAS	📈 S&P500 Multi-Factor Equity Model · 🔗 Live App Python · pandas · yfinance · Streamlit · Plotly Fully deployed multi-factor equity model across 498 S&P500 constituents (2024–2025). Implements Momentum, Value, Quality, Size, and Low Volatility with skip-1 momentum, independent cross-sectional z-scoring, and monthly quintile backtesting. Momentum (+0.83 Sharpe) and Quality (+1.17 Sharpe) were the only positive factors — consistent with the large-cap growth dominated market regime. Quantitative Finance Factor Model Backtesting Streamlit pandas
📊 Conformal Prediction Playground · 🔗 Live Demo Python · FastAPI · React · Antigravity · HuggingFace Spaces Interactive prediction intervals with finite-sample coverage guarantees. 5-agent Antigravity pipeline comparing Two-Piece Modal CP (dissertation method), OLS CP, and CQR. 3.9% efficiency gain on Wage (n=3,000). Deployed live. Conformal Prediction Agentic AI FastAPI React Statistics	👗 TempThreads AI-Powered Fashion Startup | Co-Founder Cross-platform fashion & travel assistant using multi-agent LLM pipelines (Gemini, OpenAI, Anthropic). 100+ users, 1,500+ A/B experiments, 12% accuracy gain, 18% satisfaction improvement. Flutter Firebase GCP LangChain Multi-Agent LLM
🎬 AI YouTube Shorts Generator Multi-Agent · Content Pipeline · Automation Automated pipeline for generating YouTube Shorts using AI agents for scriptwriting, visual generation, and video assembly. Python AI Agents Automation Content Generation	♟️ Chess Engine & Coaching FIDE Rated 1903 | Head Coach Competitive chess player and head coach at Wallace Chess, coaching classes of 10-25 students (ages 4-16) across London schools. Strategy Pattern Recognition Coaching

💼 Experience

🖤 BlackRock AWT (Aladdin Wealth Tech) — Technology Analyst, Client Delivery & Execution (Aug 2026 –)

Joining the Aladdin Wealth Tech team in London, working within Client Delivery & Execution on the Aladdin platform that powers risk analytics and portfolio management for wealth managers globally.

🚀 TempThreads — Co-Founder & Product Engineer

Built a cross-platform AI recommendation engine with Flutter/Firebase/GCP. Designed explainable rules-based + multi-agent systems. Ran 1,500+ controlled experiments.

🏛️ Department for Transport (GSS Fast Stream) — Data Analyst Intern

Automated the Annual Bus Operator Survey validation in BigQuery for 450+ operators. Built cosine similarity pipelines for imputation accuracy. Ensured UK Statistics Authority Code of Practice compliance.

📊 Xiatech — Data Science Intern

Power BI dashboards with BigQuery integration. Customer segmentation via k-means clustering. Regression modelling and algorithm design.

🛠️ Tech Stack

AI & ML

Statistics & Data Science

Languages

Infrastructure & Data

Frontend & Mobile

Dev Tools & Platforms

📫 Connect