• What is DEXtektif?
• Project Showcase Video
• Competition Achievement
• The Problem
• How DEXtektif Solves It
• Key Features
• Screenshots
• Repositories
• Links & Contact

DEXtektif is an AI-powered risk-monitoring platform built to help regulators and crypto businesses in Indonesia detect money-laundering activity on Decentralized Exchanges (DEX) — before it does damage. It was created for the BI-OJK Hackathon 2025.

This organization hosts the repositories that make up the DEXtektif platform — the tools regulators, compliance teams, and API consumers use to monitor wallets, investigate suspicious activity, manage blacklists, and generate compliance reports. See Repositories below for the full list.

DEXtektif was honored as a Top 15 Semifinalist of the BI-OJK Hackathon 2025, hosted by Bank Indonesia (Indonesia's Central Bank) and Otoritas Jasa Keuangan (OJK) (Indonesia's Financial Services Authority).

📌 Competition Page: https://archive.faizath.com/archive/1768636861.333245/singlefile.html (archived — the original page at https://hackathon.fekdi.co.id/ is down)

BI-OJK Hackathon 2025 is a nationwide hackathon jointly run by Bank Indonesia and OJK to source technology-driven solutions across three problem statements: AI as a Service (AIaaS) for Digital Delivered Service Export, Financial Innovation & Public Services, and Risk Management & Consumer Protection (fraud detection, AML/CFT compliance, and blockchain/on-chain analysis — the track DEXtektif competed in). It ran from its kickoff on June 5, 2025 through submissions, semifinals, and a final quarantine/judging round on October 28–29, 2025, culminating in a winner announcement and showcase at FEKDI 2025 (Oct 30–Nov 1, 2025), with prizes up to Rp 200,000,000 per category.

DeFi's openness is also its biggest blind spot. DEXtektif exists because:

• Anonymous transactions are hard to trace — DEX trades don't go through KYC'd intermediaries.
• Fake tokens are trivial to create — anyone can launch a token on-chain in minutes.
• Prices get manipulated via wash trading — inflating volume/price to lure victims.
• Regulators lack the tooling — OJK and PPATK have no purpose-built, real-time DEX monitoring system.

The scale of the problem is significant:

DEXtektif implements a two-layer hybrid AI architecture that watches Solana (the chain carrying ~60% of all crypto transactions and ~95% of memecoin trading volume) in real time:

Solana RPC Stream
       │
       ▼
Data Ingestion & Feature Engineering
       │
       ├──► Layer 1: Early Detection (Isolation Forest)
       │      → instantly scores each transaction for statistical anomalies
       │
       └──► Layer 2: Deep Analysis (Graph Neural Network)
              → maps wallet/transaction networks to uncover hidden
                collusion, wash trading & rug-pull patterns
       │
       ▼
XGBoost Aggregator → Final Risk Score (0–100)
       │
       ├──► Regulator Dashboard (OJK Intelligence)
       └──► Client API  →  { "status": "BLACKLISTED", "risk_score": 97 }

• Fast + deep, together: Layer 1 catches obviously anomalous transactions instantly; Layer 2 uncovers coordinated fraud networks that only become visible when relationships between wallets are analyzed.
• Real-time alerts: suspicious activity is flagged in under 5 minutes.
• Regulation-aligned: designed around POJK 27/2024 and FATF guidance for crypto AML/CFT.
• Built for the local gap: existing tools like Chainalysis, CipherTrace, and Elliptic are slow to flag new/small tokens and aren't tailored to Indonesian regulators — DEXtektif focuses specifically on wash trading and scam-token detection on DEXs, in near real time.

Core capabilities delivered across the DEXtektif platform:

• 🐙 GitHub: github.com/DEXtektif
• 🌐 Live Deployment: dextektif.faizath.com
• 📧 Contact: contact@dextektif.faizath.com

⚠️ Domains and the contact email above are migrating on July 24, 2026 — see the notice at the top of this README for the new addresses.