feat: added files, readme in markdown format, images, pdfs, src/ code and langflow components and flows. author: Paul Parkinson

Author: jasperan

langflow-agentic-ai-oracle-mcp-vector-nl2sql/README.md

@@ -0,0 +1,143 @@
+"""
+Oracle Database Local Embeddings Component
+
+This component provides local SentenceTransformer embeddings optimized for Oracle Database
+vector storage, ensuring consistent embedding dimensions and models.
+
+Author: Paul Parkinson
+"""
+
+from langchain_huggingface import HuggingFaceEmbeddings
+from langflow.base.models.model import LCModelComponent
+from langflow.io import DropdownInput, IntInput, BoolInput, Output
+from langflow.field_typing import Embeddings
+
+
+class OracleDatabaseEmbeddingsComponent(LCModelComponent):
+    """
+    Local SentenceTransformer embeddings optimized for Oracle Database vector storage
+    """
+
+    display_name = "Oracle Database Local Embeddings"
+    description = "Local SentenceTransformer embeddings for Oracle 23ai (384 dimensions, no cloud dependencies)"
+
+    inputs = [
+        DropdownInput(
+            name="model_name",
+            display_name="Embedding Model",
+            info="Choose the SentenceTransformer model for embeddings",
+            options=[
+                "sentence-transformers/all-MiniLM-L12-v2",
+                "sentence-transformers/all-MiniLM-L6-v2",
+                "sentence-transformers/all-mpnet-base-v2",
+                "sentence-transformers/paraphrase-MiniLM-L6-v2",
+                "sentence-transformers/distiluse-base-multilingual-cased",
+            ],
+            value="sentence-transformers/all-MiniLM-L12-v2",
+        ),
+        IntInput(
+            name="max_length",
+            display_name="Max Sequence Length",
+            info="Maximum length of input sequences",
+            value=512,
+            advanced=True,
+        ),
+        BoolInput(
+            name="normalize_embeddings",
+            display_name="Normalize Embeddings",
+            info="Whether to normalize embeddings to unit length",
+            value=True,
+            advanced=True,
+        ),
+        BoolInput(
+            name="show_progress",
+            display_name="Show Progress",
+            info="Whether to show download progress for models",
+            value=False,
+            advanced=True,
+        ),
+    ]
+
+    outputs = [
+        Output(display_name="Embeddings", name="embeddings", method="build_embeddings"),
+    ]
+
+    def build_embeddings(self) -> Embeddings:
+        """
+        Build the HuggingFace embeddings model
+        """
+        try:
+            # Configure model kwargs
+            model_kwargs = {
+                'device': 'cpu',  # Use CPU for local deployment
+            }
+
+            # Configure encode kwargs - remove show_progress_bar to avoid conflicts
+            encode_kwargs = {
+                'normalize_embeddings': self.normalize_embeddings,
+            }
+
+            embeddings = HuggingFaceEmbeddings(
+                model_name=self.model_name,
+                model_kwargs=model_kwargs,
+                encode_kwargs=encode_kwargs,
+                show_progress=self.show_progress,  # Use show_progress instead
+            )
+
+            self.status = f"✅ Local embeddings loaded: {self.model_name}"
+            return embeddings
+
+        except Exception as e:
+            error_msg = f"Failed to load embeddings model: {str(e)}"
+            self.status = f"❌ {error_msg}"
+            raise RuntimeError(error_msg)
+
+    def get_model_info(self) -> dict:
+        """
+        Get information about the selected model
+        """
+        model_info = {
+            "sentence-transformers/all-MiniLM-L12-v2": {
+                "dimensions": 384,
+                "description": "Fast and efficient, great for general purpose (recommended for Oracle DB)",
+                "size": "133MB"
+            },
+            "sentence-transformers/all-MiniLM-L6-v2": {
+                "dimensions": 384,
+                "description": "Smaller and faster version",
+                "size": "91MB"
+            },
+            "sentence-transformers/all-mpnet-base-v2": {
+                "dimensions": 768,
+                "description": "Higher quality but larger",
+                "size": "438MB"
+            },
+            "sentence-transformers/paraphrase-MiniLM-L6-v2": {
+                "dimensions": 384,
+                "description": "Optimized for paraphrase detection",
+                "size": "91MB"
+            },
+            "sentence-transformers/distiluse-base-multilingual-cased": {
+                "dimensions": 512,
+                "description": "Multilingual support",
+                "size": "540MB"
+            }
+        }
+        return model_info.get(self.model_name, {"dimensions": "Unknown"})
+
+    def validate_for_oracle_db(self) -> bool:
+        """
+        Validate that the model is suitable for Oracle Database vector storage
+        """
+        model_info = self.get_model_info()
+
+        # Oracle 23ai works best with these dimensions
+        recommended_dims = [384, 512, 768]
+        model_dims = model_info.get("dimensions", 0)
+
+        if model_dims not in recommended_dims:
+            self.status = f"⚠️ Warning: {model_dims} dimensions may not be optimal for Oracle DB"
+            return False
+
+        self.status = f"✅ Model validated: {model_dims} dimensions, Oracle DB compatible"
+        return True

langflow-agentic-ai-oracle-mcp-vector-nl2sql/components/embeddings/oracledb_embeddings.py

@@ -0,0 +1,36 @@
+
+# Oracle Database Vector Store (Langflow Custom Component)
+
+This custom component wraps `OracleVS` (Oracle 23ai Vector Store) and exposes *configurable retrieval* knobs inside Langflow.
+
+## Features
+- Local embeddings handle (connect any Embeddings node)
+- Choose distance: COSINE / EUCLIDEAN / DOT_PRODUCT
+- Search types:
+  - `similarity` (top-k)
+  - `mmr` (diversity via MMR)
+  - `similarity_score_threshold` (filter by min score; `k` becomes a cap)
+- Tunables: `number_of_results (k)`, `fetch_k`, `mmr_lambda`, `score_threshold`
+
+## Installation
+1. Save `oracledb_vectorstore.py` into your Langflow custom components folder, e.g.:
+   - macOS/Linux: `~/.langflow/components/`
+   - Windows: `%USERPROFILE%\\.langflow\\components\\`
+2. Restart Langflow. The component appears as **Oracle Database Vector Store**.
+
+## Inputs
+- **Database User / Password / DSN / Wallet Dir / Wallet Password**
+- **Table Name**: e.g., `PDFCOLLECTION`
+- **Embedding Model**: connect a local `Embeddings` node
+- **Search Query**
+- **Number of Results (k)**
+- **Search Type**
+- **Score Threshold** (only used for threshold mode)
+- **Fetch K**
+- **MMR Lambda**
+- **Distance Strategy**
+
+## Notes
+- Ensure your vector column dimensions match the embedding model dimension.
+- If using threshold mode, set a reasonable `score_threshold` (e.g., 0.3–0.4).
+- Set `fetch_k > k` for better MMR/threshold results.

langflow-agentic-ai-oracle-mcp-vector-nl2sql/components/vectorstores/README_oracle_vectorstore.md

@@ -0,0 +1,50 @@
+"""
+Local SentenceTransformer Embedding Component for Langflow
+Uses the same embedding model as the investment advisor: all-MiniLM-L12-v2
+"""
+
+from langflow.custom import Component
+from langflow.io import Output
+from langflow.schema import Data
+from langchain_community.embeddings import SentenceTransformerEmbeddings
+from typing import List
+class LocalSentenceTransformerComponent(Component):
+    display_name = "Local SentenceTransformer Embeddings"
+    description = "Local SentenceTransformer embeddings using all-MiniLM-L12-v2 (same as investment advisor)"
+    documentation = "Uses the same embedding model that created your PDF vectors"
+    icon = "🔗"
+    name = "LocalSentenceTransformerComponent"    
+    inputs = []    
+    outputs = [
+        Output(display_name="Embeddings", name="embeddings", method="build_embeddings"),
+    ]    
+    def build_embeddings(self) -> Data:
+        """Build local SentenceTransformer embeddings component"""
+        try:
+            # Use the same model as your investment advisor
+            embeddings = SentenceTransformerEmbeddings(
+                model_name="all-MiniLM-L12-v2",
+                # Cache directory to avoid re-downloading
+                cache_folder="./.cache/sentence_transformers"
+            )
+            
+            # Test the embeddings to make sure they work
+            test_embedding = embeddings.embed_query("test")
+            print(f"✅ Local embeddings working! Dimension: {len(test_embedding)}")
+            
+            self.status = f"✅ Local SentenceTransformer ready (dim: {len(test_embedding)})"
+            
+            return Data(
+                value=embeddings,
+                data={
+                    "model_name": "all-MiniLM-L12-v2",
+                    "embedding_dimension": len(test_embedding),
+                    "type": "local_sentence_transformer"
+                }
+            )
+            
+        except Exception as e:
+            error_msg = f"❌ Failed to load local embeddings: {str(e)}"
+            print(error_msg)
+            self.status = error_msg
+            raise RuntimeError(error_msg)