neel.shan_
Region: US
Friday 20 June 2025 03:24:56 GMT
78433
4065
95
286
Music
Download
Comments
Derrick :
Bros entire product is a API call to OpenAI
2025-06-20 12:37:05
1208
Samuel Groven :
just applied to join the team 🫡
2025-06-20 04:34:10
2
Oscar Olivarez :
Open ai will come out with a better version of this and blow them out the water. They are going to wish they didn’t drop out 😂😂😂
2025-06-20 17:25:01
87
Indefeasible :
crazy how this is literally the entire product
2025-06-20 04:58:31
552
Outsourced Logic :
Hire a linguistics major and teach them prompt engineering not even kidding
2025-06-21 00:28:04
58
content producer :
woahhh a chatGPT wrapper 🤯🤯
2025-06-20 14:37:09
126
jay :
Just use FAISS and compile it into Inverted Files to retrieve back old sales calls/info, instead of picking the first file, it’s faster, and better for data retrieval even for a 1% less accuracy.
2025-06-20 11:18:28
10
Violentdiarrhea :
system prompt engineer😭😭😭😭😭😭🥀
2025-06-20 09:04:00
119
Bryan 🇨🇲 :
3rd cluely video I’ve seen in under 20 mins. I don’t follow anyone associated with the company or anything. Impressive
2025-06-21 01:57:49
14
Eddie :
There was literally no reason to put that on a whiteboard 😂
2025-06-20 19:21:11
28
SG5000 :
Just asking ChatGPT to rewrite the system prompt
2025-06-21 02:19:03
11
coldestmutt :
Finally an answer to what the product is and what the company does
2025-06-20 05:48:55
28
mr.putangmaster69 :
What’s the issue with ctrl v and ctrl c into chat at this point
2025-06-20 13:56:09
3
BasicDev :
System prompts can only go so far without training your own model
2025-06-20 14:03:43
13
SD :
I think using LLms to grade llms is such a slippery slope
2025-07-24 01:46:34
1
achiexyz :
You got any recommendations on resources to learn how to setup a system prompt properly for your own usecase?
2025-06-20 08:53:09
3
hryhdegb :
Check out the judge python package. Maybe you could use diff models in the evals to get better/more insightful evals using an ensemble of LLM judges
2025-06-21 02:21:39
3
Damn:/ :
Have u tried using chatgpt ?😏
2025-06-20 06:28:32
11
RetroTidj :
Using LangGraph?
2025-06-20 20:29:35
1
Ultra :
Geeked or locked in
2025-06-21 02:01:25
1
Asher :
Something I use in Zettel is an agent dedicated to prompt injecting into the system prompt BEFORE the main agent(s) act. Basically makes any agent automatically tunable plus you only have one agent you really need to tune. Memory specific for the agent is important as well.
2025-06-20 17:40:17
2
Creek :
Try DSPy
2025-06-20 16:01:02
1
azim :
hv u used testing platforms or feedback loops right on the platform ? helped me a ton
2025-06-20 03:38:37
2
Desburryhall 🔥 :
Can you tell me how many languages did you use to implement cluely please 🙏
2025-06-20 07:18:26
2
annasweii :
Sick 🔥
2025-06-20 05:42:09
1
To see more videos from user @neel.shan_, please go to the Tikwm
homepage.
