Quant Research · Data Systems · Public Proof-of-Work

I kill-test ideas against
real data, then ship the survivors.

Quant researcher at NTU Singapore, builder of validated data and forecasting systems. I tested 13 market anomalies — 12 died. One survived: S&P 500 index additions, +2.27% per event, t-stat 2.93 across 102 events. I build the same way: validation before belief, uncertainty made explicit, everything auditable.

t-stat 2.93 · 1 of 13 signals survived
571 tests · WC2026 lab
4 World Cups leak-free validated
KEDGE → NTU → HSG St. Gallen
How I work

Proof over narrative.

Four principles run through every project. They are also why my models are smaller than they could be — I reject what I can't defend.

01 Kill-test

Every idea is assumed wrong until real data says otherwise. Most don't survive — and that's the point.

02 Validate, don't claim

Leak-free, out-of-sample, walk-forward. A model earns its place by beating a baseline, or it's cut.

03 Quantify uncertainty

Point estimates lie. I ship intervals and label what they don't cover. No false precision.

04 Make it auditable

Every number traces to a source. Reproducible offline. Tests gate the claims.

Selected work

Systems I've built & shipped.

Probabilistic forecasting● LIVE / OSS

WC2026 Forecast Lab

A live probabilistic World Cup forecast: Elo→Dixon-Coles + an ML ensemble, multi-provider live xG and odds, validated leak-free on four past World Cups, with champion-probability intervals.

571 testsML@0.204 WCs validated5 data APIs
Equity signals · EDGAR● RESEARCH

BLACK ICE

Systematic signal research on SEC EDGAR filings. Structural, forced-flow edges over behavioral noise — the S&P 500 index-addition effect ($15.6T in tracker funds) is the anchor case.

+2.27%/eventt-stat 2.93102 events
SaaS · EdTech● LIVE

Major — lemajor.fr

An all-in-one platform for French law students: AI document classification, adaptive QCM, an AI editor with correction, and unified legal search across Legifrance / EUR-Lex / CEDH.

Next.js 16SupabaseAI pipeline
Investing · automation● PRIVATE

Pro-Act Invest

A private investment-club operations system: organized data base, automated reporting pipelines, and strict confidentiality controls. Process design over hype.

automationdata ops
Quant tooling● BUILDING

AlphaEngine

A research harness for testing market anomalies end-to-end: data, hypothesis, significance testing, and an honest reject/keep gate. The infrastructure behind the kill-tests.

backtestingsignificance
Flagship case study

WC2026 — a forecast you can audit.

Most World Cup "predictions" are confident guesses. This is a probability distribution built to survive a hostile review — not to look impressive.

Data & APIs

Live data from five providers, cross-checked by a disagreement system:

  • TheStatsAPI — per-shot shotmap xG (x/y coords) + bookmaker odds
  • Highlightly — team xG / advanced stats
  • API-Football — live score, events, lineups
  • football-data.org — standings, scorers, fixtures
  • 49,450 historical matches (martj42) for Elo & validation

Model & validation

  • Calibrated Elo → Dixon-Coles Poisson core; bounded xG live adjustment
  • ML 1X2 ensemble reweights W/D/L marginals — beat the Elo baseline leak-free (Brier 0.508 vs 0.529)
  • Tournament walk-forward on WC2010/14/18/22 — retrained before each so it never cheats
  • ML weight cut 0.50 → 0.20 because the higher weight over-concentrated favorites and hurt the 2018 upset

Uncertainty & honesty

  • Champion probabilities are P5/P50/P95 intervals, not point estimates
  • Market odds used as a benchmark / control layer, not blended into the model
  • A 20-point self-hostile reviewer audit lists the project's own weaknesses

Honest caveats, stated on the project itself: the xG providers likely share an upstream (not independent); the intervals are a floor (parameter sampling only, not total uncertainty); a dynamic upset-robust ML mode exists but is not the default. It's a forecasting lab, not a calibrated guarantee.

About

Builder, not a buzzword collector.

I'm Yorian Melki, a quant researcher and systems builder. I care about the part most people skip: validation, uncertainty, and reproducibility. A model I can't defend under attack doesn't ship.

Right now I'm at NTU Business School Singapore (Simulation Techniques in Finance), building public proof-of-work across forecasting, equity-signal research, and product. The thread is the same everywhere: structure over story, evidence over confidence.

I'm heading to HSG St. Gallen next, continuing on a quant / research trajectory — and compounding a portfolio of real, auditable projects along the way.

NOW · 2026
Quant Researcher @ NTU Singapore
SEP 2026
→ HSG St. Gallen
FOUNDATION
KEDGE Business School
SHIPPING
WC2026 · BLACK ICE · Major · AlphaEngine
Contact

Let's build something defensible.

Open to quant research, data/AI systems, and serious collaboration.