Fendy Feng

Content writer & AI Builder. Love Milvus

[Fendy Feng @Fendy_like]

Love it when a product this good runs on our stack. So proud to power search for Consensus 🔥

[Zilliz @zilliz_universe]

@consensusapp , the operating system for scientific research, serves 10M+ researchers, students, and clinicians across 12,500+ universities and has answered 150M+ research questions by searching 400M+ scholarly sources.

As Consensus moved to agentic research, its search stack hit its limits:
❌ Pure keyword matching missed papers that used different wording
❌ Elasticsearch required compressing vectors to control cost, which hurt retrieval quality
❌ A full re-index took 24+ hours

The team needed search that understands meaning rather than just matching keywords, and stays fast and affordable at 400M+ vectors.

After evaluating Elasticsearch, FAISS, Pinecone, and Zilliz Cloud, Consensus chose Zilliz Cloud to power semantic retrieval behind their agentic academic search. The results:
📈 14% higher search precision
⚡ ~45 ms P99 across 400M+ vectors
🔁 Full rebuilds: 24+ hours → ~1 hour
💰 Up to 4× lower storage costs

"Zilliz Cloud gives our research agent fast, high-quality semantic retrieval, directly widening the evidence it can reach." — Christian Salem, Co-founder & CPO

We're proud that Zilliz Cloud powers semantic search for Consensus's academic research agent, and we're excited to keep building together as agentic research scales.

🔗 Read the full story: https://t.co/dA7W1Kr353

Try Zilliz Cloud now: https://t.co/gDbnx2k5Jv

[Fendy Feng @Fendy_like]

Opus 5.5 is the first Opus since 4.6 that made me go "oh, nice." 👀

4.7, 4.8, Opus 5 felt like small steps for my content work (sometimes steps back). 5.5 feels like a real writing partner again ✍️

Anyone else feel the same? 👇#opus

[Fendy Feng @Fendy_like]

Our previous webinar: ES vs. Milvus 3.0 on full-text search was a great success. We had good attendance and strong engagement. Now here comes Episode II. We'll continue to compare ES/OpenSearch vs. Milvus, plus their managed services. Save a spot now: https://t.co/ZQhVxW0eS6

[Fendy Feng @Fendy_like]

Great sharing! Thanks, Elliot. And really love Evermind's swag. Looks so pretty.👏

[Elliot 艾略特 @ElliotChen]

This is why I love meeting friends and other communities. We were lucky to be invited to the @milvusio meetup yesterday.

We shared ideas with the Milvus community and ended up on stage talking about how embedding databases work with EverOS.

People in the audience were taking photos while we talked.

Two minutes later, our friend in @evermind community had turned the moment into a bunch of memes.

This one is my favorite.

It's the Nyan Cat, of course. Now I can't get the melody out of my head.

Community hangout is the best. 💛

[Fendy Feng @Fendy_like]

@elliotchen @milvusio @yinmin1987 @louiselu007 🫡🫡🫡🫡

[Fendy Feng @Fendy_like]

We're running a webinar comparing Elasticsearch vs Milvus 3.0 on full-text search! ☺️ If you're a fan of ES, come take a look! If you love Milvus, come join us too, and you won't be disappointed. Register here: https://t.co/EedX0N92t1

[Fendy Feng @Fendy_like]

Using OpenClaw does burn Claude tokens..... 🤪 ↓ This is Claude Code Max Plan ($100)...it just reached the first token limits... 🤣

But anyway, still love it. I just can't go back to those old days without AI....

#claude #claudecode #vibecoding #openclaw

[Fendy Feng @Fendy_like]

As a marketer who can barely code, I never thought I'd build a real tool by myself. But I just #vibecoded a blog publisher with Claude Code that cut my work from 1hr to 5 min. No more manual formatting. All automated. 😃 How can I not love it? #claude
https://t.co/KtG0EW3JEl

[Fendy Feng @Fendy_like]

Cool!

[Zilliz @zilliz_universe]

Join us in celebrating the launch of Zilliz Cloud Serverless, our newest vector database solution for GenAI applications, on Product Hunt! 🎉

Check it out: https://t.co/4L6pbq97KI

Your support has been instrumental in getting us to this point. We'd love for you to join the discussion on Product Hunt and share your experience with our community.

[Milvus @milvusio]

🌟Milvus just hit 30,000 stars on GitHub!🌟

A huge thank you to our amazing community of developers and contributors who've made this possible. 🤝

#Milvus #GitHub #Vectordb #OpenSource #Community

[Milvus @milvusio]

🌟 30,000 Milvus stars 🌟 🤯 Thanks for contributing to our open-source community!

"Milvus wouldn't be the same without you - without our users, contributors, and companies using us!" @stephenbtl

#OpenSource #Milvus #Community #Milestone #GitHub #Vectordatabase

[Fendy Feng @Fendy_like]

@milvusio 🎆🎆🎆🎆🎆

[Milvus @milvusio]

Kicking off tonight's #UnstructuredData event with RAG with knowledge graph + Multimodality + #Milvus #Vectordb!

Watch it LIVE here: https://t.co/i9aKF971Sk

[Fendy Feng @Fendy_like]

New for Zilliz Cloud!!
🔄 Seamless data migration from other vector databases
🔌 Streamlined unstructured data ingestion with Fivetran
🚀 Turbo-charged query performance
⚖️ Smart resource scaling
💪 99.95% uptime SLA
🔍 Next-level system observability https://t.co/u048djfgkB

[Zilliz @zilliz_universe]

🆕 Zilliz Cloud update for #AIEngineering:

🔄 Vector DB migration
🔌 Fivetran connector (500+ sources)
🚦 Multi-replica queries
📈 Auto-scaling
🔍 Blog deep-dive: https://t.co/I42g0hA1c4

[Fendy Feng @Fendy_like]

So cute!!

[Milvus @milvusio]

The #Milvus Bird has many talents: singing, skateboarding, scuba diving, paddle boarding, etc!

New stickers coming to a meetup near you 😉
https://t.co/SrAho53Q13

[Fendy Feng @Fendy_like]

Guess how fast it could be to install and run? Seconds! And it is super light and can run on your laptop and notebook. Give it a try and start building your GenAI apps with Milvus Lite in seconds!

[Milvus @milvusio]

🎉 A new Milvus is here! Milvus Lite is an easier way to get started with vector search. Simply pip-install pymilvus on your laptop or notebooks. When you’re ready to migrate your prototype to production, the shared API makes it easy. Learn more 👉 https://t.co/HldZwszDTZ

[Fendy Feng @Fendy_like]

Great talk!

[Milvus @milvusio]

🔍 Unlock next-level search with sparse & dense embeddings!

This video dives into combining lexical search (TF-IDF, BM25) & dense models using techniques like RRF & weighted averages.

Watch the Milvus 2.4 sparse-dense vector search demo on-demand: https://t.co/Hnk7loW4Ie

[Jiang Chen @jiangc1010]

Milvus is now available as a retrieval module (MilvusRM) in DSPy! DSPy is a programmatic LLM optimization framework by @stanfordnlp, providing composable and declarative modules for instructing LMs in Pythonic syntax. Compared to prompting through trial and error, it can take query and answer examples as input and acts as an agent to search for a prompt that meets the expectations. It can also use tools such as MilvusRM to conduct retrieval augmentation (RAG).