refactor/recommender-modular-structure#64

Open

DYung26 wants to merge 53 commits into

PlotSenseAI:developfrom

DYung26:refactor/recommender-modular-structure

Commits on Oct 30, 2025

feat(core): implement `AIModelInterface` for unified multi-provider `LLM` querying with strategy-based execution, parallel model handling, multimodal message building, and weighted model scoring
DYung26
committed
feat(utils): add `API` key prompt handling, plot saving, and image encoding functions for `matplotlib` integration
DYung26
committed
feat(enums): add `StrategyName` enum with round-robin, cost-optimized, performance, and fallback options
DYung26
committed
feat(providers): add `AnthropicProvider` for `Claude` model integration with key validation and model listing
DYung26
committed
feat(providers): add `AzureOpenAIProvider` for `Azure`-hosted `OpenAI` models with endpoint, key validation, and model listing
DYung26
committed
feat(providers): define abstract `LLMProvider` base class with query, list_models, and validate_key methods
DYung26
committed
feat(providers): add `GeminiProvider` for `Google Gemini` models supporting text and multimodal queries with validation
DYung26
committed
feat(providers): add `GroqProvider` for fast Groq API access with key validation and supported models listing
DYung26
committed
feat(providers): add `GroqProvider` via `OpenAI`-compatible `SDK` endpoint with query, list_models, and key validation
DYung26
committed
feat(providers): add `OllamaProvider` for local `OpenAI`-compatible `Ollama` models with query and validation
DYung26
committed
feat(providers): add `OpenAIChatProvider` for `OpenAI Chat` models with query, list_models, and key validation
DYung26
committed
feat(providers): add `OpenAIResponseProvider` for `OpenAI Responses API` with query, list_models, and key validation
DYung26
committed
feat(providers): add `ProviderManager` to orchestrate multiple `LLM` providers, manage `API` keys, and handle queries
DYung26
committed
feat(strategies): add `CostOptimizedStrategy` to prioritize cheaper models with iteration-based escalation
DYung26
committed
feat(strategies): add `FallbackChainStrategy` to try models in fixed order until success, with optional success tracking
DYung26
committed
feat(strategies): add `PerformanceOptimizedStrategy` to prioritize highest-performance models with fallback to lower tiers
DYung26
committed
feat(strategies): add `RoundRobinStrategy` to evenly cycle through all models for balanced usage
DYung26
committed
feat(strategies): define abstract `Strategy` base class for selecting provider/model pairs with iteration-aware selection
DYung26
committed
refactor(explanations): rewrite `PlotExplainer` to use `AIModelInterface` and `ProviderManager` with strategy-based model selection, remove `Groq`-specific hardcoding, and unify image handling
DYung26
committed
style(init): fix import spacing for consistency and add trailing newline to comply with `PEP8` formatting
DYung26
committed
build(deps): expand requirements with pinned dependency versions and add new `AI/LLM` packages; update `setup.py` to read `README` safely and include additional dependencies
DYung26
committed
chore(egg-info): update plotsense metadata, dependencies, authorship, license, and examples to `v0.1.3`
DYung26
committed

Commits on Oct 31, 2025