krzysztof/landing_page
1
0
Fork 0
mirror of https://github.com/qdrant/landing_page.git synced 2026-10-04 18:38:30 +02:00
Code Issues Packages Projects Releases Wiki Activity

renaming files for compatibility

Browse Source
This commit is contained in:
sabrinaaquino
2025-04-15 13:06:00 -03:00
parent 6f2ecd40ef
commit 73c2a8fcce
5 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
qdrant-landing/content/blog/case-study-lyzr.md
+4 -4
View File
@@ -18,7 +18,7 @@ tags:
---
# How Lyzr Supercharged AI Agent Performance with Qdrant
![How Lyzr Supercharged AI Agent Performance with Qdrant](/blog/case-study-lyzr/case-study-lyzr-summary-dark.jpg)
![How Lyzr Supercharged AI Agent Performance with Qdrant](/blog/case-study-lyzr/case-study-lyzr-summary-dark.png)
## Scaling Intelligent Agents: How Lyzr Supercharged Performance with Qdrant
@@ -30,7 +30,7 @@ This is how they rethought their stack and adopted Qdrant as the foundation for
## The Scaling Limits of Early Stack Choices
![Lyzr-architecture](/blog/case-study-lyzr/lyzr-architecture.png)
![Lyzr-architecture](/blog/case-study-lyzr/lyzr-architecture.jpg)
Lyzr’s architecture used Weaviate, with additional benchmarking on Pinecone. Initially, this setup was fine for development and controlled testing. The system managed around 1,500 vector entries, with a small number of agents issuing moderate query loads in a steady pattern.
@@ -97,7 +97,7 @@ One deployment, built for NTT Data, focused on automating IT change request work
After migrating to Qdrant, the difference was immediate. Retrieval accuracy improved substantially, even for long-tail queries. The system maintained high responsiveness under concurrent loads, and horizontal scaling became simpler—ensuring consistent performance as project demands evolved.
![NTT Architecture](/blog/case-study-lyzr/NTT-visual.png)
![NTT Architecture](/blog/case-study-lyzr/ntt-visual.jpg)
---
@@ -107,7 +107,7 @@ Another example involved NPD, which deployed customer-facing agents across six w
Qdrant’s vector search enabled accurate, low-latency retrieval across thousands of entries. Even under increasing user traffic, the platform delivered consistent performance, eliminating the latency spikes experienced with previous solutions.
![NPD Architecture](/blog/case-study-lyzr/NPD-visual.png)
![NPD Architecture](/blog/case-study-lyzr/npd-visual.jpeg)
## Final Thoughts