adaptive-intelligence 0.1.0__py3-none-any.whl

adaptive_intelligence/__init__.py

@@ -0,0 +1,23 @@
+"""
+Adaptive Intelligence — Self-Improving Retrieval Orchestration Framework
+
+Drop documents. Ask questions. The system learns how to retrieve better over time.
+
+Usage:
+    from adaptive_intelligence import AdaptiveAI
+
+    engine = AdaptiveAI()
+    engine.ingest("./documents")
+    response = engine.ask("What are the operational risks?")
+    print(response.answer)
+    print(response.confidence)
+    print(response.citations)
+"""
+
+__version__ = "0.1.0"
+__author__ = "VK Venkatkumar"
+
+from adaptive_intelligence.core.engine import AdaptiveAI
+from adaptive_intelligence.core.response import AdaptiveResponse
+
+__all__ = ["AdaptiveAI", "AdaptiveResponse"]

adaptive_intelligence/core/__init__.py

@@ -0,0 +1,3 @@
+from adaptive_intelligence.core.engine import AdaptiveAI
+from adaptive_intelligence.core.response import AdaptiveResponse
+from adaptive_intelligence.core.config import AdaptiveConfig

adaptive_intelligence/core/config.py

@@ -0,0 +1,149 @@
+"""Configuration for Adaptive Intelligence engine."""
+
+from dataclasses import dataclass, field
+from typing import Optional, Dict, Any, List
+from enum import Enum
+from pathlib import Path
+
+
+class LLMBackend(str, Enum):
+    OLLAMA = "ollama"
+    OPENAI = "openai"
+    AZURE_OPENAI = "azure_openai"
+    ANTHROPIC = "anthropic"
+    HUGGINGFACE = "huggingface"
+    GROQ = "groq"
+    TOGETHER = "together"
+    CUSTOM = "custom"
+
+
+class SecurityLevel(str, Enum):
+    STANDARD = "standard"
+    HIGH = "high"
+    MAXIMUM = "maximum"
+
+
+class Domain(str, Enum):
+    GENERAL = "general"
+    FINANCIAL = "financial"
+    LEGAL = "legal"
+    TECHNICAL = "technical"
+    HEALTHCARE = "healthcare"
+    SCIENTIFIC = "scientific"
+    OPERATIONAL = "operational"
+
+
+@dataclass
+class ChunkingConfig:
+    chunk_size: int = 500
+    chunk_overlap: int = 50
+    respect_boundaries: bool = True
+    min_chunk_size: int = 100
+
+
+@dataclass
+class RLConfig:
+    warmup_queries: int = 15
+    exploration_rate: float = 0.2
+    min_exploration_rate: float = 0.05
+    exploration_decay: float = 0.995
+    learning_rate: float = 0.1
+    discount_factor: float = 0.95
+    update_frequency: int = 5
+    algorithm: str = "thompson_sampling"  # "epsilon_greedy", "ucb", "thompson_sampling"
+
+
+@dataclass
+class GraphConfig:
+    enabled: bool = True
+    conditional_activation: bool = True
+    max_hops: int = 3
+    min_entity_count_for_activation: int = 2
+    relationship_confidence_threshold: float = 0.5
+    activation_success_threshold: float = 0.7
+
+
+@dataclass
+class EvaluationConfig:
+    faithfulness_weight: float = 0.30
+    relevance_weight: float = 0.20
+    citation_weight: float = 0.20
+    precision_weight: float = 0.10
+    recall_weight: float = 0.10
+    latency_penalty_weight: float = 0.05
+    token_cost_weight: float = 0.03
+    hallucination_penalty_weight: float = 0.02
+    llm_judge_threshold: float = 0.8
+    enable_llm_judge: bool = True
+    enable_consistency_checks: bool = False
+
+
+@dataclass
+class AdaptiveConfig:
+    # LLM Configuration
+    llm_backend: LLMBackend = LLMBackend.OLLAMA
+    llm_model: str = "llama3.2"
+    base_url: Optional[str] = None
+    api_key: Optional[str] = None
+    azure_endpoint: Optional[str] = None
+    deployment_name: Optional[str] = None
+    temperature: float = 0.1
+    max_tokens: int = 2048
+
+    # Domain
+    domain: Domain = Domain.GENERAL
+
+    # Security
+    security_level: SecurityLevel = SecurityLevel.STANDARD
+    network_enabled: bool = False
+    enable_audit_trail: bool = True
+
+    # Storage
+    storage_dir: str = "./.adaptive_intelligence"
+    vector_db: str = "chromadb"
+
+    # Chunking
+    chunking: ChunkingConfig = field(default_factory=ChunkingConfig)
+
+    # RL Engine
+    rl: RLConfig = field(default_factory=RLConfig)
+
+    # Graph
+    graph: GraphConfig = field(default_factory=GraphConfig)
+
+    # Evaluation
+    evaluation: EvaluationConfig = field(default_factory=EvaluationConfig)
+
+    # Retrieval
+    default_retrieval_depth: int = 5
+    max_retrieval_depth: int = 20
+    enable_reranking: bool = True
+
+    # Logging
+    log_level: str = "INFO"
+    log_file: Optional[str] = None
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Serialize config to dictionary."""
+        import dataclasses
+        return dataclasses.asdict(self)
+
+    @classmethod
+    def from_dict(cls, data: Dict[str, Any]) -> "AdaptiveConfig":
+        """Create config from dictionary."""
+        # Handle nested dataclasses
+        if "chunking" in data and isinstance(data["chunking"], dict):
+            data["chunking"] = ChunkingConfig(**data["chunking"])
+        if "rl" in data and isinstance(data["rl"], dict):
+            data["rl"] = RLConfig(**data["rl"])
+        if "graph" in data and isinstance(data["graph"], dict):
+            data["graph"] = GraphConfig(**data["graph"])
+        if "evaluation" in data and isinstance(data["evaluation"], dict):
+            data["evaluation"] = EvaluationConfig(**data["evaluation"])
+        if "llm_backend" in data and isinstance(data["llm_backend"], str):
+            data["llm_backend"] = LLMBackend(data["llm_backend"])
+        if "security_level" in data and isinstance(data["security_level"], str):
+            data["security_level"] = SecurityLevel(data["security_level"])
+        if "domain" in data and isinstance(data["domain"], str):
+            data["domain"] = Domain(data["domain"])
+        return cls(**data)