Sage Nord Lab

Research packages

• Research package Mobile/API NIS2 Technical Gap Check 25,000-35,000 DKK · 3-5 business days A fixed-fee technical review of your mobile app and API surface against NIS2's Article 21 style requirements, with a written gap report and executive summary. Not a compliance audit, not a certification. View package →

Articles

• PureReason: a deterministic verification layer that catches AI hallucinations in under 5ms September 17, 2026 · 4 min read Most hallucination detection asks a second LLM to grade the first one, which is slow, expensive, and only as reliable as the grader. PureReason is Sage Nord's open-source alternative: symbolic logic and embeddings instead of another model, verifying AI output in under 5 milliseconds with zero API calls. Read article →
• Cortyx: why your coding agent shouldn't re-read the whole repo on every turn September 17, 2026 · 5 min read Long-lived AI coding sessions keep paying to re-establish context: either the full repo gets stuffed into every prompt, or a memory layer calls another LLM to decide what matters. Cortyx is Sage Nord's open-source answer: a local, pure-Rust context delivery engine that retrieves the right 3-5 pre-written 'neurons' in under 40ms, with no runtime model and 96.8% R@5 on LME-500. Read article →
• Why Uber built one gateway for 60+ internal LLM use cases instead of letting every team integrate separately September 5, 2026 · 3 min read Uber had over 60 internal LLM use cases with each team building its own integration. The GenAI Gateway centralized vendor access, security, and cost tracking behind an OpenAI-compatible API, and it also shows what that centralization costs. Read article →
• How Discord fixed hot partitions by moving trillions of messages off Cassandra September 5, 2026 · 3 min read Discord's Cassandra cluster grew to 177 nodes and kept paging on-call for unpredictable latency caused by 'hot partitions.' The fix wasn't just a database swap to ScyllaDB, it was a coalescing layer in front of it. Read article →
• How Shopify sharded Shop app's Rails backend without stopping the world August 31, 2026 · 3 min read Shopify's engineering team migrated the Shop app's MySQL backend to Vitess to fix multi-week schema migrations and capacity throttling. The real lesson isn't the sharding key, it's the query verifiers that made the migration safe. Read article →
• Why on-device LLM assistants have felt slow, and how NPUs fix it August 31, 2026 · 3 min read A Peking University ASPLOS 2025 paper shows NPU offloading can cut mobile LLM prefill latency dramatically. Here's why that matters for anyone building on-device AI features on iOS or Android. Read article →
• EU AI Act Article 50 is now in force: what chatbots, generative AI, and deepfake features must do August 31, 2026 · 3 min read Article 50 of the EU AI Act took effect on 2 August 2026, requiring AI transparency disclosures and machine-readable watermarking. Here's what it actually requires, verified against the official text, and what it means for teams shipping AI features. Read article →
• What 86,726 broken code samples tell us about trusting AI-generated code August 11, 2026 · 3 min read A new IEEE ISSRE 2026 study quantifies how LLM-generated code actually fails across languages, and shows that self-correction is far less reliable than assumed. Here's what it means for teams shipping AI-assisted code. Read article →

From our authors