📈 AdVantage

AI-powered advertising intelligence and budget optimization — maximize ROAS across campaigns with ML-driven budget allocation, creative performance prediction, and automated bidding across Google, Meta, TikTok, and DV360.

ROI Dashboard • Quick Start • Architecture • API • Modules • Contributing

---

⭐ Maximizing ad ROI? Star AdVantage to support open-source ad optimization!

---

💰 ROI Dashboard

  CAMPAIGN           SPEND      REVENUE    ROAS    STATUS          ACTION
  ──────────────────────────────────────────────────────────────────────
  Q3 Brand          $45,200    $187,300   4.14x   ● Optimal        ─
  Q3 Performance    $32,800    $98,400    3.00x   ● Optimal        ─
  Q3 Retarget       $18,500    $24,050    1.30x   ⚡ Optimize     Reallocate $5K↗
  Q3 Awareness      $12,000    $9,600     0.80x   ⚠️ Review        Pause or revise
  ──────────────────────────────────────────────────────────────────────
  TOTAL             $108,500   $319,350   2.94x                      ▲ 18% vs Q2

Features

Feature	Description	Impact
Budget Optimizer	Convex optimization + multi-armed bandit allocation	+22% ROAS on avg
Creative Scoring	Predicts ad CTR and conversion rate before launch	3.4× better creative selection
Automated Bidding	Real-time bid adjustments based on conversion probability	31% CPA reduction
Agent Framework	Autonomous agents monitor campaigns and suggest reallocations	24/7 optimization
Forecasting	Time-series spend and revenue predictions with uncertainty bounds	±8% MAPE at 30 days
Multi-channel	Google Ads, Meta, TikTok, DV360, The Trade Desk	Unified platform

---

Quick Start

# Install Go API server
go install github.com/Crynge/AdVantage/src/api/server

# Start optimization API
advantage-server --port 8080 --config config.yaml

from advantage.optimizer import BudgetOptimizer

optimizer = BudgetOptimizer(
    budget=100000,
    channels=[
        {"name": "google", "roas": 3.8, "volatility": 0.2},
        {"name": "meta", "roas": 2.9, "volatility": 0.3},
        {"name": "tiktok", "roas": 4.2, "volatility": 0.5},
        {"name": "dv360", "roas": 2.1, "volatility": 0.4},
    ],
)

allocation = optimizer.optimize()
# {'google': 42000, 'meta': 31000, 'tiktok': 27000, 'dv360': 0}

---

Architecture

flowchart TB
    subgraph Sources["Data Sources"]
        G[Google Ads API] --> I[Ingestion Layer]
        M[Meta API] --> I
        T[TikTok API] --> I
        D[DV360 API] --> I
    end

    subgraph Engine["Optimization Engine"]
        I --> C[Cube: Cache / Store]
        C --> O[Budget Optimizer]
        C --> S[Creative Scorer]
        C --> B[Bid Controller]

        O --> AG[Agent Monitor]
        S --> AG
        B --> AG
    end

    subgraph API["API Layer"]
        AG --> API_Server[Go REST API]
        API_Server --> U[Web Dashboard]
        API_Server --> CLI[CLI Tool]
    end

    subgraph Actions["Actions"]
        AG --> REC[Recommendations]
        AG --> ALERT[Alerts]
        AG --> AUTO[Auto-Adjust]
    end

Loading

---

API

# Run budget optimization
curl -X POST http://localhost:8080/api/optimize \
  -H "Content-Type: application/json" \
  -d '{"budget": 100000, "channels": ["google", "meta", "tiktok"]}'

# Get campaign list
curl http://localhost:8080/api/campaigns

# Generate forecast
curl -X POST http://localhost:8080/api/forecast \
  -d '{"campaign_id": "q3-brand", "horizon_days": 90}'

from advantage.agents import CampaignMonitor

monitor = CampaignMonitor(channels=["google", "meta"])
while True:
    for alert in monitor.check():
        print(f"[{alert.severity}] {alert.channel}: {alert.message}")
    await asyncio.sleep(3600)

---

Modules

src/
├── api/
│   └── server.go              # Go REST API
├── advantage/
│   └── optimizer.py           # Budget optimization engine
└── agents/
    └── agent.py               # Autonomous campaign agents

---

Contributing

See CONTRIBUTING.md for guidelines.

• Open an issue

---

License

MIT

---

🌐 Crynge Ecosystem

All repos are free and open-source. ⭐ Star what you use!

Category	Repos
LLM & AI	SpecInferKit · AetherAgents · PromptShield
Marketing	AdVerify · Attributor · InfluencerHub · EdgePersona · AdVantage · BrandMuse · CampaignForge
Simulation	CivSim · EvalScope
Operations	OpsFlow

_{Built by Crynge · ⭐ Star us on GitHub!}