vector search
technique judged 98% likely a named technique or architecture
See how it has spread across media tiers
Mentions over time
Table view
| Week | Articles | Sources |
|---|---|---|
| 2022-W28 | 1 | 1 |
| 2023-W51 | 1 | 1 |
| 2024-W05 | 1 | 1 |
| 2024-W08 | 1 | 1 |
| 2024-W11 | 2 | 2 |
| 2024-W13 | 1 | 1 |
| 2024-W15 | 3 | 1 |
| 2024-W22 | 1 | 1 |
| 2024-W23 | 1 | 1 |
| 2024-W24 | 1 | 1 |
| 2024-W25 | 1 | 1 |
| 2024-W26 | 1 | 1 |
| 2024-W29 | 1 | 1 |
| 2024-W33 | 1 | 1 |
| 2024-W40 | 1 | 1 |
| 2024-W41 | 1 | 1 |
| 2024-W47 | 3 | 2 |
| 2024-W49 | 1 | 1 |
| 2024-W50 | 1 | 1 |
| 2024-W52 | 1 | 1 |
| 2025-W02 | 1 | 1 |
| 2025-W09 | 1 | 1 |
| 2025-W21 | 1 | 1 |
| 2025-W22 | 1 | 1 |
| 2025-W23 | 1 | 1 |
| 2025-W24 | 1 | 1 |
| 2025-W27 | 1 | 1 |
| 2025-W35 | 1 | 1 |
| 2025-W40 | 1 | 1 |
| 2025-W49 | 1 | 1 |
| 2026-W01 | 1 | 1 |
| 2026-W04 | 1 | 1 |
| 2026-W07 | 1 | 1 |
| 2026-W08 | 1 | 1 |
| 2026-W09 | 4 | 1 |
| 2026-W11 | 12 | 2 |
| 2026-W12 | 1 | 1 |
| 2026-W13 | 3 | 2 |
| 2026-W15 | 1 | 1 |
| 2026-W16 | 1 | 1 |
| 2026-W18 | 1 | 1 |
| 2026-W19 | 1 | 1 |
| 2026-W20 | 1 | 1 |
| 2026-W24 | 4 | 2 |
| 2026-W25 | 1 | 1 |
| 2026-W32 | 3 | 3 |
| 2026-W33 | 1 | 1 |
| 2026-W34 | 2 | 1 |
| 2026-W35 | 1 | 1 |
| 2026-W36 | 3 | 2 |
| 2026-W37 | 2 | 1 |
| 2026-W38 | 3 | 1 |
| 2026-W39 | 2 | 2 |
| 2026-W40 | 19 | 5 |
| 2026-W41 | 6 | 2 |
In their words
Structure your context retrieval as intent-specific pipelines: Intent: urgent_delivery ├─ Primary: Order status (DynamoDB, <100ms) ├─ Secondary: Shipping partner API (network call, <500ms) ├─ Tertiary: Customer tier/VIP status (cache, <50ms) └─ Fallback: Previous support tickets (search index) Intent: refund_request ├─ Primary: Transaction record (database) ├─ Secondary: Refund policy version (con
Return only a number between 0 and 1.""" response = await self.judge.ainvoke(prompt) try: scores.append(float(response.strip())) except: scores.append(0.5) return sum(scores) / len(scores) if scores else 0 def _compute_consistency(self, all_outputs) -> float: # Semantic similarity between runs (simplified) agreements = 0 total = 0 for outputs in all_outputs: for i in range(len(outputs)): for j in
A plain similarity search gives you the closest chunks.
embedder ( query ) # Stage 2: Vector search (fetch more for reranking) raw_results = self .
split () return sentence_bleu ( ref_tokens , pred_tokens ) def compute_bertscore ( references : list [ str ], predictions : list [ str ]) -> dict : """ BERTScore: semantic similarity using embeddings.
Extend it with: Database tools (SQL queries) API integrations (Slack, GitHub, Notion) Browser automation (Playwright) Image generation (DALL-E, Stable Diffusion) Vector search (your RAG system!) The pattern is always: define tool → register schema → execute → return result.
Coordination Patterns, Version-Aware | Dimension | LangGraph | CrewAI | AutoGen (existing systems) | Microsoft Agent Framework (new projects) | |---|---|---|---| | Status | Actively developed | Actively developed | Maintenance mode — no new features; community-managed | 1.0 GA April 2026 ; Microsoft's recommended successor | | Mental model | Explicit graph: nodes are steps, edges are transitions |
Strategy How it works Best for Watch out for Fixed-size Split every N tokens, usually with overlap Quick baselines, uniform text Cuts mid-sentence and mid-idea Recursive Split on paragraphs, then sentences, then words, until chunks fit General prose; a solid default Ignores document structure beyond separators Structure-aware Split on headings, sections, code blocks, table boundaries Docs, wikis,