Agent skill · data analytics · jeremylongshore
apollo-performance-tuning
Optimize Apollo.io API performance.
Why this skill is useful
Adds executable scripts for optimizing Apollo.io API performance through caching, connection pooling, and bulk operations that are not commonly known.
What it needs
Requires node installed locally. Requires apollo account access. About 4k tokens when loaded. Last updated 2026-08-07. 2,604 stars on the source repository.
What this skill does
Apollo Performance Tuning Overview Optimize Apollo.io API performance through response caching, connection pooling, bulk operations, parallel fetching, and result slimming. Key insight: search is free but slow (~500ms), enrichment costs credits — cache aggressively and batch enrichment calls. Prerequisites Valid Apollo API key Node.js 18+ Instructions Step 1: Connection Pooling Reuse TCP connections to avoid TLS handshake overhead on every request. Step 2: Response Caching with Per-Endpoint TTLs Step 3: Use Bulk Endpoints Over Single Calls Apollo's bulk enrichment endpoint handles 10 records per call vs 1. Massive performance gain. Step 4: Parallel Search with Concurrency Control Step 5: Slim Response Payloads Apollo returns large person objects (~2KB each). Extract only needed fields to reduce memory. Step 6: Benchmark Your Endpoints Output Connection pooling with keepAlive and configurable maxSockets LRU cache with per-endpoint TTLs (24h org, 4h contact, 15m search) Bulk enrichment via /people/bulkmatch (10x fewer requests) Parallel search with p-queue concurrency control Response slimming reducing memory from ~2KB to ~200B per person Benchmarking script measuring avg and p95 latency Examples For a search-latency regression, capture a baseline with the free search and health endpoints against a mock or approved sandbox fixture, then introduce connection reuse and a bounded cache behind a feature flag. Compare p95, error rate, cache-hit behavior, and result equivalence before promoting the change. A paid-enrichment benchmark requires an explicit approval flag, a fixed credit budget, and a public organization fixture; otherwise it remains out of the run. Roll back the flag if latency improves at the cost of stale, incorrect, or over-broadly cached results. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills jeremylongshore/apollo-performance-tuning