| As AI gets more capable, its answers still depend on what it can find.
Vespa Voice is back for Season 2 ๐๏ธ Our CEO andโฆ |
@vespaengine |
Company |
Original |
2026-11-04 |
180 |
3 |
1 |
0 |
0 |
0 |
| ๐ช๐ต๐ฎ๐ ๐ถ๐ ๐ฎ๐ป ๐๐ ๐ฆ๐ฒ๐ฎ๐ฟ๐ฐ๐ต ๐ฃ๐น๐ฎ๐๐ณ๐ผ๐ฟ๐บ? ๐๐ฒ๐ ๐๐ต๐ฒ ๐ผ๐๐ฒ๐ฟ๐๐ถ๐ฒ๐ ๐ถ๐ป ๐ท๐๐๐ ๐ผ๐๐ฒ๐ฟ ๐ฎ ๐บ๐ถ๐ป๐๐๐ฒ. ๐ง
AI applications need to find relevant informโฆ |
@vespaengine |
Company |
Original |
2026-11-03 |
190 |
3 |
0 |
0 |
0 |
2 |
| ๐๐
๐๐ฒ๐ป๐ฑ๐ถ๐ป๐ด ๐ฉ๐ฒ๐๐ฝ๐ฎโ๐ ๐๐๐ฏ๐ฟ๐ถ๐ฑ ๐ฆ๐ฒ๐ฎ๐ฟ๐ฐ๐ต ๐๐ถ๐๐ต ๐ช๐ผ๐ฟ๐บ๐ต๐ผ๐น๐ฒ ๐ฉ๐ฒ๐ฐ๐๐ผ๐ฟ๐ ๐
Most hybrid search runs lexical and semantic search separatelyโฆ |
@vespaengine |
Company |
Original |
2026-10-29 |
172 |
5 |
0 |
1 |
0 |
1 |
| How do you modernize recommendations when your data is anything but perfect?
At https://t.co/abkb8IknIf Live London, Dโฆ |
@vespaengine |
Company |
Original |
2026-10-27 |
171 |
4 |
0 |
0 |
0 |
1 |
| How Etsy replaced Solr and several other retrieval engines with one Vespa.ai-based system
Gregg Donovan shared the lesโฆ |
@vespaengine |
Company |
Original |
2026-10-22 |
208 |
3 |
1 |
0 |
0 |
0 |
| Building the foundations for AI search
At https://t.co/abkb8IjPSH Live London, Cem รnsal of Walmart drew on roles at โฆ |
@vespaengine |
Company |
Original |
2026-10-21 |
165 |
1 |
0 |
0 |
0 |
0 |
| Quepid now works with Vespa!
Eric Pugh from OpenSource Connections recorded a short demo of Vespa integrated into Quepโฆ |
@vespaengine |
Company |
Original |
2026-10-20 |
262 |
5 |
0 |
0 |
0 |
1 |
| RT @thenewstack: No reranker can recover a relevant passage that retrieval failed to surface. Bonnie Chase and Jenny Moโฆ |
@vespaengine |
Company |
Repost |
2026-10-09 |
21 |
0 |
1 |
0 |
0 |
0 |
| Thomson Reuters chose https://t.co/fQ26wwi1By
Read the blog post explaining their reasoning and experience here: httpsโฆ |
@vespaengine |
Company |
Original |
2026-10-05 |
576 |
10 |
5 |
0 |
0 |
2 |
| RT @thenewstack: What happens when the right answer never makes it into the candidate set? Join Bonnie Chase and Jenny โฆ |
@vespaengine |
Company |
Repost |
2026-09-30 |
18 |
0 |
1 |
0 |
0 |
0 |
| RT @thenewstack: Human-scale retrieval and agent-scale retrieval are two very different infrastructure problems. Whit Wโฆ |
@vespaengine |
Company |
Repost |
2026-09-16 |
23 |
0 |
1 |
0 |
0 |
0 |
| Whoโs https://t.co/abkb8IjPSH Live for? If youโre building search, RAG, personalization, or recommendation systems at sโฆ |
@vespaengine |
Company |
Original |
2026-06-25 |
285 |
8 |
0 |
2 |
0 |
1 |
| A nice overview of the features that makes Vespa the AI search platform https://t.co/fb9mvuxNce |
@vespaengine |
Company |
Original |
2026-06-22 |
299 |
4 |
0 |
1 |
0 |
2 |
| Image search usually means stitching together a vector database, a keyword engine, and an inference service: three systโฆ |
@vespaengine |
Company |
Original |
2026-06-22 |
3,599 |
27 |
8 |
7 |
0 |
14 |
| Stop tuning your vector search. Start budgeting it โฑ๏ธ Vespa's HNSW search can now terminate early - and it changes how โฆ |
@vespaengine |
Company |
Original |
2026-06-16 |
1,569 |
11 |
2 |
2 |
0 |
3 |
| Vespa isn't only powering American AI, now it's also @MistralAI's retrieval solution https://t.co/GpS7bFZZUL |
@vespaengine |
Company |
Original |
2026-05-29 |
518 |
6 |
2 |
1 |
0 |
0 |
| The May https://t.co/abkb8IjPSH Newsletter is out!
This month weโre announcing updates focused on retrieval quality, rโฆ |
@vespaengine |
Company |
Original |
2026-05-28 |
871 |
9 |
2 |
8 |
1 |
0 |
| Vinted explains how to get stable p99 latency of 3ms when using Vespa as a feature store. ๐ https://t.co/Q3Pk55IYl9 |
@vespaengine |
Company |
Original |
2026-04-13 |
667 |
7 |
0 |
2 |
0 |
4 |
| AWS has published a sample app which shows how to build a scalable LLM agent grounded in both general and proprietary iโฆ |
@vespaengine |
Company |
Original |
2026-04-13 |
427 |
5 |
0 |
1 |
0 |
0 |
| Itโs happening. https://t.co/abkb8IjPSH is going live, in person.
For the first time ever, weโre bringing the Vespa coโฆ |
@vespaengine |
Company |
Original |
2026-04-07 |
297 |
7 |
3 |
1 |
0 |
1 |
| Big models just got easier to use in Vespa.
Many state-of-the-art embedding models, like Multilingual-E5-large or Embeโฆ |
@vespaengine |
Company |
Original |
2026-04-07 |
406 |
5 |
1 |
1 |
0 |
0 |
| If your application has more queries than writes, you can save a lot by using a small embedding model running locally oโฆ |
@vespaengine |
Company |
Original |
2026-03-27 |
460 |
7 |
0 |
1 |
0 |
0 |
| Will Vespa be killed by Claude?๐คhttps://t.co/fleaxM1A93 |
@vespaengine |
Company |
Original |
2026-03-20 |
250 |
4 |
0 |
0 |
0 |
0 |
| Filters are everywhere in real-world vector search, and they're quietly killing your recall and latency.
Join Radu Gheโฆ |
@vespaengine |
Company |
Original |
2026-03-18 |
343 |
9 |
1 |
0 |
0 |
1 |
| Everyone is talking about AI agents.
But the hard part isnโt the agent. Itโs retrieval.
Agents need to search, filter,โฆ |
@vespaengine |
Company |
Original |
2026-03-10 |
276 |
11 |
1 |
0 |
0 |
1 |
| Choosing the right embedding model for search.
A hard problem because this impacts memory and cpu/gpu usage, latency,โฆ |
@vespaengine |
Company |
Original |
2026-01-14 |
483 |
8 |
1 |
1 |
1 |
3 |
| The December Vespa newsletter is out.
๐ขNew features shipped:
- Automated ANN performance tuning
- Accelerated vector dโฆ |
@vespaengine |
Company |
Original |
2025-12-05 |
2,014 |
13 |
2 |
1 |
1 |
0 |
| Du you understand tensors?
๐The advent of tensors starts today:
- A new fun puzzle every day.
- Solutions are posted tโฆ |
@vespaengine |
Company |
Original |
2025-12-01 |
519 |
6 |
2 |
1 |
2 |
1 |
| When you use Vespa instead of spending your days on plumbing https://t.co/umAMqZZ9Tv |
@vespaengine |
Company |
Original |
2025-11-25 |
749 |
11 |
0 |
1 |
0 |
2 |
