Ep162: Improving Search for Generative AI Developers with DataStax and AWS
カートのアイテムが多すぎます
ご購入は五十タイトルがカートに入っている場合のみです。
カートに追加できませんでした。
しばらく経ってから再度お試しください。
ウィッシュリストに追加できませんでした。
しばらく経ってから再度お試しください。
ほしい物リストの削除に失敗しました。
しばらく経ってから再度お試しください。
ポッドキャストのフォローに失敗しました
ポッドキャストのフォロー解除に失敗しました
-
ナレーター:
-
著者:
このコンテンツについて
Learn how DataStax transformed customer feedback into a hybrid search solution that powers Fortune 500 companies through their partnership with AWS.
Topics Include:
- AWS and DataStax discuss how quality data powers AI workloads and applications.
- DataStax built on Apache Cassandra powers Starbucks, Netflix, and Uber at scale.
- Their TIL app collects outside-in customer feedback to drive product development decisions.
- Hybrid search and BM25 kept trending in customer requests for several months.
- Customers wanted to go beyond pure vector search, not specifically BM25 itself.
- Research showed hybrid search improves accuracy up to 40% over single methods.
- ML-based re-rankers substantially outperform score-based ones despite added latency and cost.
- DataStax repositioned their product as a knowledge layer above the data layer.
- Developer-first design prioritizes simple interfaces and eliminates manual data modeling headaches.
- Hybrid search API uses simple dollar-sign parameters and integrates with Langflow automatically.
- AWS PrivateLink ensures security while Graviton processors boost efficiency and tenant density.
- Graviton reduced total platform operating costs by 20-30% with higher throughput.
Participants:
- Alejandro Cantarero – Field CTO, AI, DataStax
- Ruskin Dantra - Senior ISV Solution Architect, AWS, Amazon Web Services
See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/
まだレビューはありません