Apoorva Mishra

Product Thinker | Vibe Coder | HBR Author | Creator of - ;

[Apoorva Mishra @apswin]

@romanbuildsaas GTM

[Apoorva Mishra @apswin]

@MichLieben ads

[Apoorva Mishra @apswin]

@gregisenberg @gregisenberg If anyone on the internet who is doing more than just giving motivation and real practical tools to win with AI, its you.

[Apoorva Mishra @apswin]

@garrytan @bscholl Love it, if anything, AI is eating software engineering as its making everyone get into code, coding was not as fun before this

[Apoorva Mishra @apswin]

SDKs are back, APIs / MCPs are intermediate product layer - but Agentic UX is still getting evolved, who solves it first, sets the precedent

[Apoorva Mishra @apswin]

@lukepierceops PLAYBOOK

[Apoorva Mishra @apswin]

@KevinSzabo14 They solve the right problem

[Apoorva Mishra @apswin]

@pmitu How do you recommend to start? How can we automate and test our markting
pipeline?

[Apoorva Mishra @apswin]

@pmitu I think its mostly long form AI slop

[Apoorva Mishra @apswin]

@jesse_vermeulen doomscrolling

[Apoorva Mishra @apswin]

@claudeai In a way this was already possible but more structured, might be a tough road ahead for Lovable

[Apoorva Mishra @apswin]

Speed is moat

[Gumroad @gumroad]

Start ugly, improve later. Your customers will thank you for not waiting.

[Apoorva Mishra @apswin]

@pmitu How do you recommend to start? How can we automate and test our marketing
pipeline?

[Apoorva Mishra @apswin]

@ravikiran_dev7 Great point, what would they be, may be headless desktop apps for AI agents

[Apoorva Mishra @apswin]

@dramaricic Handles everything = true, fragmented yes and that is the opportunity, not everyone is capable of agentic workflow automation

[Apoorva Mishra @apswin]

@aryanlabde Time to be one, what are your tips

[Apoorva Mishra @apswin]

@pmitu Expand on #2 please

[Apoorva Mishra @apswin]

@jackmoses777 This is perfect for openclaw automation workflow

[Apoorva Mishra @apswin]

@daniel_dhawan Would love to check this out

[Apoorva Mishra @apswin]

@shl What are the best ways to prevent AI from converging to statistically learned generic slop? How to lean it away from that?