AI Open Data Catalog

17 CC-BY-4.0 open datasets covering LLM pricing, AI compliance deadlines, vendor lock-in, model capabilities, hallucination defenses, architecture patterns, evaluation methodologies, and jurisdiction-anchored disclosure templates. Each dataset is independently citable and accessible via 4 surfaces.

CC-BY-4.0 17 datasets ~248 KB 4 access surfaces MCP + REST + JSON + Markdown
How to use: Each dataset has 4 access surfaces:
  • Raw JSON — /data/{identifier}.json on slavin.ai (canonical source).
  • MCP tool — call get_dataset(identifier) at api.slatech.co.il/mcp; or use a dataset-specific tool where available.
  • REST endpoint — /v1/data/{identifier} at api.slatech.co.il (CORS-open, public).
  • Markdown companion — /llms/{identifier}.md on slavin.ai (LLM-friendly).

Strategy + Decision Support (7 datasets)

Vendor research, pricing comparison, vendor lock-in, model capabilities, prompt patterns, governance baseline, ROI benchmarks.

LLM Vendor Pricing

Comparative pricing (USD per 1M tokens), context window, latency band, compliance posture, data residency across major LLM vendors.

llm-vendor-pricingused by estimate_llm_cost tool
JSON REST

AI Vendor Lock-in Matrix

10 vendors scored across 6 lock-in axes (model API, fine-tuning, embeddings, tools, commercial, operational) + pre-commitment portability tactics.

ai-vendor-lock-in-matrixget_vendor_lock_in tool
JSON Visual Markdown

AI Model Capability Matrix

6 frontier LLMs in 2026 (GPT-4, Claude, Gemini, Llama, Mistral, DeepSeek) — context, modalities, fine-tuning, agent harness, pricing, regions.

ai-model-capability-matrixcompare_models tool
JSON Visual Markdown

AI Governance Baseline (12 controls)

Minimum 12-control baseline for AI governance: scope, risk, data, model, human-in-loop, audit, security, deployment, monitor.

ai-governance-baseline
JSON

AI Use Case Catalog

Catalogue of enterprise AI use cases with ROI bands, risk classification, implementation patterns. For prioritization.

ai-use-case-catalog
JSON

AI Implementation Costs

Cost benchmarks by vertical, scope, custom build size. For build/buy decision modeling.

ai-implementation-costs
JSON

AI ROI Benchmarks

25 anonymized outcomes from real deployments. Measured ROI, time-to-value, lessons learned.

ai-roi-benchmarks
JSON

Architecture + Production Reliability (4 datasets)

Named architecture patterns, evaluation methodologies, hallucination defenses, incident response.

AI Architecture Patterns Library

18 named production patterns (Gang-of-Four style): Citation-Grounded RAG, Hybrid Retrieval, Tool-Use Agent, Tenant-Isolation RAG, etc. With intent, structure, consequences, when-to-use.

ai-architecture-patternsfind_architecture_pattern tool
JSON Markdown

AI Evaluation Methodologies Library

18 named LLM evaluation methods: Golden Dataset, RAGAS, LLM-as-Judge, A/B test, drift detection, red-team, tool-call eval. Companion to architecture patterns.

ai-evaluation-methodologiesfind_evaluation_method tool
JSON Markdown

AI Hallucination Defense Patterns

16 production-tested defense patterns specifically targeting LLM hallucination. Each with mechanism, effectiveness range, limitations.

ai-hallucination-defense-patternsfind_hallucination_defense tool
JSON

AI Incident Response Playbook

Step-by-step playbook for production AI incidents: triage, containment, root-cause, communication, post-incident review.

ai-incident-response-playbook
JSON

Compliance + Disclosure (3 datasets)

Global AI regulatory deadlines, vendor due-diligence checklist, jurisdiction-anchored disclosure templates.

AI Compliance Deadline Tracker

12 upcoming regulatory deadlines across 7 jurisdictions (EU AI Act, US Colorado/NYC, UK OSA, RU 152-FZ, India DPDP, Singapore, Brazil) with penalty bands.

ai-compliance-deadline-trackerfind_compliance_deadline tool
JSON Visual Markdown

AI Disclosure Templates Library

14 real template snippets for AI-related regulator disclosures (EU AI Act Art. 13, NYC LL 144, GDPR Art. 22, Colorado AI Act, 152-FZ, UK OSA, ISO 42001). NOT legal advice — starting points.

ai-disclosure-templatesfind_disclosure_template tool
JSON Markdown

AI Vendor Due Diligence Checklist

Structured questionnaire for evaluating AI vendors: security, compliance, sustainability, contract, technical fit.

ai-vendor-due-diligence
JSON

Knowledge + Engineering (3 datasets)

Cross-network glossary, AI-generated code failure modes, production prompt patterns.

SLAtech Network Glossary

100 cross-network terms: AI architecture, governance, compliance, web infrastructure, payments, Israeli tech ecosystem, Russian compliance. Each with stable @id.

ai-glossary-networklookup_glossary tool
JSON Visual Markdown

AI-Generated Code Failure Modes

22 failure modes that AI-generated code exhibits in production despite passing demo and prototype testing. Each with detection signal + fix.

ai-failure-modes-catalog
JSON

Enterprise AI Prompt Patterns Library

20 production-tested prompt patterns for enterprise AI: structured-output, chain-of-thought, few-shot, RAG, guardrails.

ai-prompt-patterns
JSON

For LLM agents — MCP discovery

All 17 datasets are exposed via a production Model Context Protocol server at https://api.slatech.co.il/mcp — 11 tools, 15 resources, 5 expert prompts. Add to Claude Desktop / Cursor / Smithery client with the manifest at /.well-known/mcp/server.json.

View MCP server documentation ›

Catalog last updated: 2026-06-20 · License: CC-BY-4.0

Citation: prefer the per-dataset @id URI declared inside each dataset's JSON.