~ヾ(＾∇＾)
Quantum Computing · Tensor Networks · World Models · Quantitative Finance
Quantum computing engineer working on tensor network methods, JEPA/SIGReg-based world models, and their application to finance, sports analytics, and physical systems.
- PhD, Quantum Computing — IIT Bombay (quantum-inspired tensor network optimization)
- MScFE — WorldQuant University (stochastic modeling, derivative pricing, financial econometrics, deep learning for finance)
- Co-Founder, Head of AI & Quantum Computing — Alethic, Lisbon (belief-state world models for finance, football, energy, and healthcare)
- Co-Founder, Head of Quantum — Rune Technology, Lisbon (quantum-mathematical frameworks for market intelligence)
- IBM Qiskit Advocate
- Previously Senior Engineer, Advanced Digital Technology at Hero MotoCorp
Core
Infra & Data
Applied
- World models: JEPA/SIGReg pretraining, attention-pooling, cross-domain encoders (finance, ECG, exoplanet transit, football)
- Tensor networks: MPS, MPO, Tucker and CP decomposition; quantum simulation
- Quantitative finance: signal generation, regime detection, factor modeling, live paper trading infrastructure
- Reinforcement learning: offline RL (TD3+BC), two-level PPO, multi-objective and constrained optimization
- Quantum ML: variational quantum circuits as classifier heads, quantum-inspired active inference
- ML systems: production pipelines, GCP-based automation, model validation, real-time inference
- Vantage — real-time football tactical AI predicting defensive collapses and shot timing seconds ahead of play, built for Alethic's Futebol vertical; demoed live to investors
- RUNE-Q — quantum-inspired active inference system modeling the global economy as an open quantum system (density-matrix state, Lindblad evolution, quantum MCTS)
- Shared JEPA/SIGReg world-model core validated and ported across three domains: financial regime encoding (live paper trading), exoplanet transit classification (MAESTRO), and 12-lead ECG diagnostics (CADENCE)
- Live paper trading system on Alpaca across multiple strategies, with a JEPA-encoded regime signal feeding a two-level PPO trading policy
- Quantum tensor pipeline for FX, covering 25 currency pairs and 8 geographic regions over a 26-year panel
- Automated ISO 26262 test-generation pipeline (multi-agent LangGraph) for automotive functional safety
- Belief-state world models as an alternative to purely generative LLMs for domains requiring falsifiable predictions
- Tensor network and quantum methods applied beyond simulation — real-world engineering and financial systems
- Regime detection and crisis prediction in complex adaptive systems
- Where and why variational quantum circuits under- or out-perform classical baselines under matched capacity
- Cross-border scaling of early-stage deep tech ventures
- Collaboration on world models, quantum ML, quantum simulation, and tensor network applications
- Technical discussion on JEPA-style pretraining, RL theory, quantitative finance methodology
- Deep tech / fintech / sports-analytics opportunities at this intersection
Note: Much of my recent work involves proprietary systems and confidential research. Happy to discuss methodology and approach within appropriate boundaries.