Yiftach Shoolman retweeted
🪦 Redis: IMMORTAL
Death Score: 8/100
"Congrats, you can't replace 'runs in 0.3ms at the HW level' with an MD file."
Cause of death: N/A. You cannot kill what responds before the request is sent.
Are YOU just a .md file? deathbyclawd.com?url=redis.i…
Amazing!
Redis is the second fastest growing platform - AI agents and RAG are big drivers. Kudos to @Redisinc community who are building Redis for AI.
I guess being the fastest Vector Database in the world makes a difference
Yiftach Shoolman retweeted
With Chain Of Thought reasoning (ala openAI o1-preview): intelligence now scales with inference compute
Which will cause inference costs to rise significantly
Which makes what we are doing at Redis with inference caching / semantic caching so critical to control your AI costs
Yiftach Shoolman retweeted
More amazing tech built on Redis. Paradigm makes data collection fast and easy at any scale. It uses an army of AI agents to collect data and organize it into spreadsheets in seconds, all using natural language prompts. Sign up and see more here: paradigmai.com/join
Introducing Paradigm – a reimagined workspace with AI at its core.
Centered around the primitive of a spreadsheet, Paradigm puts swarms of intelligent agents at your fingertips.
The real power of Paradigm comes with scale: imagine having tens of thousands of interns working for you in parallel.
Paradigm is 1000x faster than manual data collection, completing an average of 500 cells per minute.
Check out Paradigm in action below.
#Redis Cloud is now on #Vercel! Enjoy Redis at speed with the latest data structures, query engine, & full-text search. Plus, Vercel users can now harness Redis for AI:
1️⃣ Fastest vector search 2️⃣ Semantic Caching (save 30%+ on GenAI) 3️⃣ LLM Memory & personalization 4️⃣ Semantic routing 5️⃣ Agentic memory 6️⃣ Feature Store for enriched inferencing
#AI #RedisCloud #GenAI @Redisinc redis.io/blog/redis-cloud-no…
#Redis is evolving to power the GenAI-based modern stack!
1. Caching ➔ Semantic Caching
2. Session Management ➔ LLM Session Management (Contextual Memory)
3. Classification ➔ Semantic Routing
@Redisinc #GenAI #ModernStack
github.com/redis/redis-vl-py…
#Redis 7.4 is out!
1. Hash Field Expiration (most requested feature ever!)
2. New vector data types: less memory, higher QPS, lower latency
3. Easier secondary indexing: no special TAG chars, empty field support, new geo search (intersect/disjoint)
4. Time series size reduction with insertion filters
Check it out: redis.io/blog/announcing-red…
(@Redisinc )
Check out our blog for impressive results: #Redis #vector search is up to x50 faster than general-purpose DBs and x15 faster when combining throughput & latency, than special purpose vector DBs and imitators . Essential for #GenAI apps!
redis.io/blog/announcing-fas…
@Redisinc
Real-time #RAG apps need more than traditional vector DBs. #Redis offers:
Vector search up to 50x faster (benchmarks soon)
Semantic cache reducing LLM calls by 30%
Advanced LLM memory for personalized and more accurate responses
Designed for speed, scalability, accuracy, and cost. Details here:
redis.io/blog/using-redis-fo…
(@Redisinc )
One of the most requested features from the #redis community is now available as a release candidate - Hash Fields expiration.
( lnkd.in/eMpE9r_G ).
Furthermore, we added new bfloat16 and float16 vector data types in #Redis Stack, reducing memory consumption while maintaining accuracy.
( lnkd.in/eSv4Ka6N )
We are building Redis for real-time data and AI. More soon...
(@Redisinc)
Semantic Kernel's Kernel Memory now uses #Redis! - another great and straightforward integration that helps LLM applications run faster.
redis.io/blog/building-llm-a…
(@Redisinc)
Yiftach Shoolman retweeted
Today, the bright minds of engineers and technologies prevailed. In democracies, they thrive 🇮🇱🇺🇸
המוחות המבריקים של המהנדסים והטכנולוגיות ניצחו היום. בדמוקרטיות הם משגשגים 🇮🇱🇺🇸
In the #Intel Vision 2024 keynote, @PGelsinger (@intel's CEO) showed how RAG architecture should look with Gaudi 2 and #Redis, powered by vector search and semantic caching.
The bottom line: much more accurate and relevant results @ x2.3 times faster.
youtube.com/watch?v=QB7FoIpx…
As part of our efforts to keep #redis awesome, we made Lettuce, the wonderful Java client, an official family member.
Thank you, Mark Paluch, for making Lettuce what it is now after 10 years of hard work!
redis.com/blog/lettuce-joins…
@Redisinc
It's been a big week for #redis (and @Redisinc )! After yesterday's license change, we announced today the acquisition of #Speedb, the market's fastest storage engine. We're excited to welcome Speedb team! And here is how we see the future of #redis: redis.com/blog/the-future-of…
After 11 incredible years as @Redisinc’s CTO, I'm thrilled to step into the role of President. I'll remain deeply involved in shaping our strategy and executing our tech vision, while also engaging more with our customers and peers.
Together with our customers, partners and our great community members, we're poised to take #redis to even greater heights! 🚀
#Redis Vector Library streamlines GenAI development with an efficient client interface & Redis Enterprise boosts performance with real-time vector database capabilities for vector search, LLM caching, & chat history management. @redisinc #GenerativeAI
redis.com/blog/introducing-t…
TL;DR from my interview with AIM:
1. #Redis has been downloaded over 12 million times daily (up from 3.8 million).
2. #Redis Enterprise is used by more than 0.5 billion Indians (in the interview, half a million is written, but it is half a BILLION).
3. Country visitors to @Redisinc websites (.io - developers, .com - commercial): #1 US, #2 China, #3 India. The top city is Bangalore, and the second is Mumbai.
Lastly, we discussed the importance of Redis vector search, invented before GenAI existed.
analyticsindiamag.com/how-re…