Jeddah, Saudi Arabia · Available for hire

Match intelligence.

MNG × CS2 ANALYST

CS2 Match Intelligence — drop a demo, get HLTV-accurate stats, AI coach insights, heatmaps and scouting in seconds. Built for performance staff.

CRAFT

Human · Directed

POWER

AI · Amplified

SOUL

Saudi · First

scroll

Competitive Intelligence

CS2 match analysis.

Drop a CS2 demo — the engine parses it, computes HLTV-accurate stats, detects key moments, maps death zones, and writes a deep English coach report. Click any match below to read the full analysis.

HLTV-accurate K/D & ADRDeep tactical breakdownsHeatmaps & zones

DROP .DEM FILE

CS2 demo · HLTV / FACEIT

CHECKING BACKEND…

Match Overview—

Match Breakdown

Player PerformanceRATING* ↓

Player	K	D	ADR	KAST	RATING*

* Rating is approximated by awpy and differs slightly from HLTV. K/D, ADR, and utility counts are exact.

Utility Usage

Death Heatmap

Key MomentsAUTO

Rnd	Player	Type	Tick

Map Zones · DeathsHOTSPOTS

Entry FraggingOPENING DUELS

Player	Entry K	Entry D	Success

Team Intelligence

Team reports.

Season-level profiles built by the MNG engine across every analyzed map: WPA-anchored FragScore, five-dimension role profiles, and team tactical signals. Every rate carries a 95% Wilson confidence interval, and the written report is grounded against the data — no invented numbers.

FragScore · WPA-anchoredWilson 95% CIGrounded AI report

Capabilities · The engine

Built on real data.

A full analytical pipeline — deterministic, reproducible, grounded in real match data. Every rate carries a confidence interval; every number traces back to the tick it came from.

01 / 06

Win Probability

An empirical round-state model — every situation scored from real outcomes, smoothed with Laplace, made monotonic via PAVA. No guesswork, just the data.

Empirical · Wilson CI · PAVA

02 / 06

Anti-Strat Engine

Opponent tendencies surfaced with confidence: bombsite leans, utility timings, default patterns — each rate bounded by a Wilson interval so small samples don't lie.

Tendencies · Utility · Geography

03 / 06

VOD Review Accelerator

Every kill scored by win-probability swing, then ranked into a timestamped clip index — so review starts at the rounds that actually decided the map.

WPA per kill · Clip index

04 / 06

Player Development

Role profiles and a WPA-anchored FragScore that separate impact from raw frags — who opens, who trades, who closes rounds, measured against real leverage.

Role profiles · FragScore

05 / 06

Tactical Recommendations

Suggestions derived from real round outcomes, not opinion — where to take map control, when to commit utility, which sites to pressure against this opponent.

Derived · Outcome-based

06 / 06

Validated Event Store

Built on an atomic-fact event store spanning real tournament maps — every metric reproducible from the same source, every claim grounded, nothing hallucinated.

93 maps · 7 tournaments

About · MNG

Most people prompt AI.
I engineer it.

I built a CS2 match-intelligence engine from scratch — it breaks down demos with HLTV-grade accuracy, finds the moments that decide rounds, and writes a coach-ready report. Every number traces back to the data.

Based in Jeddah, KSA — currently looking for the right team. Built in Jeddah, built to compete. Open to esports analyst and performance-staff roles.

Match intelligence.

CS2 match analysis.

Team reports.

Built on real data.

Hire me. Or just say hi.

Hire me.
Or just say hi.