rajit

7 ms·

1.
by rajit 26d ago
why would someone use this tool over Codex? Can't I use Codex with any model?
2.
by rajit 3mo ago
It is possible to do it (ex: run a background process that analyzes memories across customers and updates the system prompt based on the findings). A specific implementation would depend on your application. Feel free to email me rajit@pris
3.
by rajit 3mo ago
Yes by using us developers are deferring the harness engineering onto us, and they can spend time writing code for their business logic. We are closer to infrastructure than a library or framework; we give developers a live agent they can c
4.
by rajit 3mo ago
Good question. There a few differences between our approach and shipping an agent with the Claude agent sdk. 1. Our approach has cron-based or trigger-based automations built-in. Building automations with claude agent sdk requires setting u
5.
by rajit 3mo ago
Developers with customer-facing chat products are the ideal customer. If a startup has a specific flow they want the agent to take and their traffic is bursty, then I'd recommend using a framework like Mastra and deploying onto a sandb
6.
by rajit 3mo ago
> what are the cost and security implications? Cost is the token usage and container uptime. > One Docker container per-customer sounds like it would be really expensive. The advantage is per-user memory and self-learning. For context
7.
by rajit 3mo ago
The most valuable pieces of information an AI agent startup can gather is access to their customer's proprietary data and knowledge of their customers preferences (memory + self-learning). Even as the cost of writing code goes to zero,
8.
by rajit 3mo ago
Thanks for the feedback. The main idea is that today to built a best-in-class agent, developers build the agent loop, session management, tools, memory, skills, automations (cron + trigger-based), sandboxed deployment, and self-learning. By
9.
Building agents without harness engineering(rajitkhanna.com)
29 points by rajit 3mo ago|16 comments
10.
by rajit 3mo ago
when will the graph memory layer be available?
11.
by rajit 6mo ago
Yes, email me at rajit@prismvideos.com, and I can connect you. Their pricing page is here https://fal.ai/pricing .
12.
by rajit 6mo ago
Prism and Higgsfield are both similar in that we bring many AI models into one place. Higgsfield is focused on a number of different use cases - storyboarding, ai filmmaking, and visual effects - while Prism is hyper-focused on short form v
13.
by rajit 6mo ago
We support Kling 3.0 on our platform, similar to Higgsfield. You can see some presets here: https://prismvideos.com/workspace/templates .
14.
by rajit 6mo ago
This is a great point, and I agree with you. If a weight loss supplement brand were to use an AI influencer to market their product, it does raise questions about whether their supplement does in fact work on real people. Nevertheless, thin
15.
by rajit 6mo ago
This is a great point. It is challenging to know which models are good at what. We've found that Seedance is good at photorealitic faces, Kling is fantastic at generating audio (highest quality model in terms of syncing character'
16.
by rajit 6mo ago
We access models through Fal ( https://fal.ai ). We offered day 0 support for Kling 3.0 and launch models on our platform the day they are live. Would be curious to see your script.
17.
by rajit 6mo ago
How do you identify "wrong tool" invocations (how is the "wrong tool" defined)?
18.
by rajit 11mo ago
We spoke to a number of browser agent companies who said deterministic RPA with an AI fallback was their "secret" :)
19.
Show HN: Code Compass – Coding agent that finds the best task to work on next(codecompass.sh)
2 points by rajit 11mo ago|0 comments
20.
by rajit 1y ago
This is a great point. I'm assuming when you mention blast radius you're mentioning the risk of the account being banned for being a bot. One risk with these new standards for agent auth - which we will of course support if our cu
21.
by rajit 1y ago
Feel free to reach out. I'm rajit [at] prismai [dot] sh.
22.
by rajit 1y ago
We setup an agent mailbox with Agentmail ( https://agentmail.to/ ). Whoever owns the account (likely the developer) sets up a forwarding rule to this account. When our agent signs in, we input the forwarded otp code to get ac
23.
by rajit 1y ago
We enable stealth mode on Kernel (onkernel.com), which is a great piece of infrastructure.
24.
by rajit 1y ago
This is a great question. We are broadly interested in the agent access problem (which in some cases may involve solving Captchas). Right now, we're focused on building connectors for our customers, which has not yet involved Captcha s
25.
by rajit 1y ago
It's a pleasure to work with you. Excited to expand to more login cases and support login to more websites.
26.
Show HN: Prism – Let browser agents access any app(prismai.sh)
21 points by rajit 1y ago|15 comments
27.
by rajit 1y ago
this is so powerful - MCP on steroids :). what are the next use cases you're looking to build?