Manik Aggarwal (@manikagg01) 's Twitter Profile
Manik Aggarwal

@manikagg01

Building the personal memory layer for LLMs | CORE (@TheRedplanethq)
Open-source: github.com/RedPlanetHQ/co…
YC alum

ID: 1913492065

linkhttps://heysol.ai/core calendar_today28-09-2013 07:34:04

364 Tweet

246 Takipçi

945 Takip Edilen

Manik Aggarwal (@manikagg01) 's Twitter Profile Photo

AI tools were supposed to save us time! Last week: → Brainstormed in ChatGPT (10 min explaining context) → Opened Cursor to build (10 min explaining again) The irony? A 2025 study found devs tasks took 19% LONGER with AI tools.Why? We lose 5+ hours weekly just re-explaining

Manik Aggarwal (@manikagg01) 's Twitter Profile Photo

Most founders overthink niche selection. I definitely did. During YC, Dalton gave us advice that sounds simple but was powerful: Pick the path where you're most confident you can get someone to try your product faster. At CORE, we were torn between: • B2B companies building

Most founders overthink niche selection. I definitely did.

During YC, Dalton gave us advice that sounds simple but was powerful:

Pick the path where you're most confident you can get someone to try your product faster.

At CORE, we were torn between:
  • B2B companies building
Manik Aggarwal (@manikagg01) 's Twitter Profile Photo

AI memory tools are getting it backwards. Vector DBs give you facts: "User is founder" + "Pivoted to AI" + "Worked on PM tools" Knowledge graphs give you the story: "User pivoted from PM to AI BECAUSE of context loss" One is similarity. The other is meaning. Knowledge graphs