Measure the cost of an AI workflow
Compare retrieval choices using cost per useful answer, source quality and latency.
Ideas, explained.
The OriginChainDB Journal.
Practical guides to databases, search, and AI.
From first principles to working applications.
In focus
How vector search works: HNSW explained
Embeddings turn meaning into geometry. A plain-language walk through similarity metrics, why brute force dies at scale, HNSW, and what quantization costs.
Read the storyKeep exploring
RSS feedWindow functions, correlated subqueries
OriginChainDB's SQL surface now has ROW_NUMBER, RANK, DENSE_RANK, LAG, LEAD and aggregate OVER, plus correlated EXISTS, IN and scalar subqueries.
Automatic Idempotency-Key in the SDK
An implementation note on automatic Idempotency-Key headers, bounded deduplication caches and retry behaviour. Check the SDK version and endpoint contract.
Write backpressure: 429 + Retry-After
OriginChainDB refuses overload with HTTP 429 and a Retry-After computed from the token bucket's refill schedule. The per-API-key limits, and retry code.
Per-key TTL: agent memory that forgets
Per-key TTL lets each record expire on its own: a ttl in seconds on any write, invisible to reads at expiry, compacted in the background. Maximum 365 days.
Idempotent tool calls for agent loops
Use stable operation IDs, stored results and reconciliation to handle retries without assuming external actions are atomic.