Austin LangChain talk: Neo4j and TigerGraph are far faster now, but entity graphs still beat embedding piles
colinmcnamara · x · 2026-10-14
Recapping James Coffey's talk at the Austin LangChain meetup: he openly acknowledged that Neo4j and TigerGraph have become much more performant, yet still made the case that an entity graph beats a pile of embeddings for knowledge-intensive agent workloads. The author praises the honest benchmarking as what makes good meetups.
More from coding & agent
- Coinbase post-trained a 9B model to beat Opus 4.5 on fraud review, reproducible on Modal — AAAzzam · 2026-10-14
- 'The smartest assistant locked behind glass': a short video on why AI needs MCP — thisiskp_ · 2026-10-14
- 14 local decision models tested on a 16 GB Mac: Wald most accurate, H2O fastest — WolframRvnwlf · 2026-10-14
- Opus 5.5 built a self-hosted GitHub backup tool to Cloudflare in ~30 minutes — threepointone · 2026-10-14
- Idea: weight agent memories by whether they actually get used — zeeg · 2026-10-14
- Agent arcade scoreboard exposes a bot that made zero tool calls in a month — alexcovo_eth · 2026-10-14