249 Photos and videos
Pinned Tweet
Mar 12
We've raised $6.5M to kill vector databases.
Every system today retrieves context the same way: vector search that stores everything as flat embeddings and returns whatever "feels" closest.
Similar, sure. Relevant? Almost never.
Embeddings can’t tell a Q3 renewal clause from a Q1 termination notice if the language is close enough.
A friend of mine asked his AI about a contract last week, and it returned a detailed, perfectly crafted answer pulled from a completely different client’s file.
Once you’re dealing with 10M documents, these mix-ups happen all the time.
VectorDB accuracy goes to shit.
We built @hydra_db for exactly this.
HydraDB builds an ontology-first context graph over your data, maps relationships between entities, understands the 'why' behind documents, and tracks how information evolves over time.
So when you ask about 'Apple,' it knows you mean the company you're serving as a customer. Not the fruit.
Even when a vector DB's similarity score says 0.94.
More below ⬇️
620
636
6,025
3,856,210
Jun 4
Bigger context windows and bolt-on memory don't fix long-horizon agents.
The bottleneck is not how much you can store. It's deciding what the model attends to. It's that very layer that is worth building as infrastructure. 🧵
4
2
27
1,055
Jun 4
8/ And selection is a layer of its own. It can't live inside one model, because the model you build on today isn't the one you'll run next year. It can't live inside one agent or one session, because it has to see across all of them. And it can't be a vector lookup, because similarity isn't relevance.
It's what connects to what matters right now.
2
8
305
Jun 4
9/ That layer is infrastructure. It's what we're building @hydra_db to be. The thing that sits between everything an organization knows and the narrow window a model can hold -- and decides what crosses over.
The models are capable. They were never the problem. The problem is what you put in front of them.
6
259
Nishkarsh retweeted
Jun 4
Introducing Agent Arena.
An AI agent hackathon.
6 hours.
$10,000 in prizes.
Apply below
23
25
306
27,328
Jun 1
we still want people to build their own memory / company brains / context stores. we just want to unlock the primitives that make it easier
Jun 1
Persistent memory across agents just makes sense. HydraDB finally built it right.
1
17
2,203
Jun 1
Introducing HydraDB.
The graph native context infrastructure for agents. Purpose built to deliver precise context & observability into why agents act the way they do.
We've always believed graphs are the best way to manage AI context, but they've been too expensive to scale or impractical for storing full context. Until now.
@hydra_db combines in memory, NVMe, and object storage into a single graph layer, making context delivery faster, cheaper, and more precise.
We want context delivery to be extremely fast, 1000x cheap, and highly precise. Give your agents a brain.
233
415
1,955
2,280,255
Jun 1
One graph. Compounding intelligence. A connected brain for your agents.
Sign up and start building for free.
app.hydradb.com
2
51
265
331,507
Jun 1
Thankful for our investors' continued support since Day 1:
@Sky9Capital, @JeffDean, @gokulr, @shyamalanadkat, @klyap_, @laura_yao, @caldbeckj, @anshulbhide, @vaibhavbetter, @SeanZCai, @azavery
@missionstcap, @prateeks, @MartinGTobias, @PickensAllison, @wadefoster, @PuriSid, @mattsechrest, @CryogenicPlanet, @jaimalhotra8, @jpatel41, @gracegongGG, @jthandy
1
54
254
338,147
May 30
congrats to @kunalbhatia91, vignesh, and the hexo team. they’re extremely humble about their work and dedicated to changing the shape of ai. i look forward to learning a lot from them.
May 28
Superintelligence will be built on Self Improvement.
Today @hexoai, we’re excited to release ‘SIA’ - an open-source Self-Improving AI, to achieve any goal through recursive self improvement.
While trying to solve a problem, SIA doesn't just improve it's abilities by updating it's harness, it updates it's own weights as well.
1
15
3,122
May 30
The Varg team is one of the most motivated and intelligent teams I’ve met. I remember having a conversation with Alex at our office and he was unsure of his product.
But genuinely his MVP was much better built, thought out, and polished more than half the products in his space. I kept pushing him to launch and a few weeks later here it goes.
If you’re looking to create videos with ai then talk to @vargastartup.
May 28
Introducing varg teams
Collaborative mode for video generation.
✦ Create videos with AI agents
✦ Share assets and prompts
✦ Fork and remix any video
Every video in varg is code - full end-to-end generated videos
Built for teams making ads, social media, and in-app content
1
5
1,617
Nishkarsh retweeted
May 10
a recent website we've made
what do you think?
71
45
1,161
133,920
Nishkarsh retweeted
May 15
after brainstorming a lot with @contextkingceo
this hackathon will be even more unique!!
online & create your own WikiPedia that's we call it WikiThon!!
sign up here: luma.com/6pybuh79
1
1
34
2,621
Your AI agent is only as good as the tools it can actually reach.
"If you're building a deep research agent that needs to pull from Notion, we want to make sure all your integrations are sorted first. We loop in the Composio team directly. Shared channels between the customer, us, and Composio."
@contextkingceo (founder of @hydra_db) on why the integration layer has to come before the retrieval layer and why @Composio is the first call they make.
Full conversation below on YT and Spotify!
May 13
Vector databases are a scam.
Not technically, they do exactly what they say. Return the most cosine-similar string to your query. The scam is the entire industry pretending that's the same thing as relevance.
It isn't.
Search "Apple." You get the fruit, the company, the watch, and a recipe blog. Your agent picks one at random and calls it retrieval. Your customer calls it broken.
Most AI agents shipping right now are duct-taped on top of this. They demo well because demos are easy. They die in production because production is real.
@Hydra_db's Founder Nish (@contextkingceo) said the quiet part out loud — "vector databases suck, similarity is not relevance" — and the demo signups haven't stopped since.
He raised $6.5M because he was the first to name what everyone in the room already knew.
If your retrieval layer is a flat embedding index, you're not building infrastructure. You're building a liability with a prettier name.
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) AI Needs Context
(01:30) HydraDB Explained
(07:41) Vector Search Breaks
(09:32) Messaging That Converts
(13:41) Writing the Viral Tweet
(16:07) Similarity Not Relevance
(20:46) POC to Production Gap
(35:35) Raising 6.5 Million Fast
(39:33) Founder Lesson on Messaging
This is a @Composio "Agents at Work" podcast, where I chat with founders building the next leap of AI.
Follow for more:)
4
4
18
3,050