Will I fail if I am building a GenAI startup at the application layer?
Almost 99% of the GenAI startups in India I speak to play at the top of the stack. It is also considered the least defensible area.
(There are a couple I have spoken to in the AI tools and Security space – which is very interesting, but a topic for another day)
There are 5000+ AI tools that allow you to design your own courses to create unique sites – check <thereisanaiforthat [dot] com.
What you can do, another can replace – hence you should go vertically down the stack could be an answer
Buts it’s a shallow response.
Here is how I tend to think about it – Ask yourself if :
1️⃣ If you are building a JTBD (Jobs to be done) that can be replaced by an existing LLM capability.
2️⃣ If not, is the JBTD exhaustive in solving for a pain the ‘customer desperately wants’ that may extend to capabilities beyond GenAI?
In all honestly, GenAI isn’t a magic wand and you don’t need to force it into your stack – but are you building features that actually help solve a point – GenAI or Otherwise?
3️⃣ Are you leveraging the foundational models but integrating it with human context + proprietary data?
My weekend project is to build a thesis around what are investable opportunities across the stack – if you have come across players who you think are doing a great job, pls DM
#ScalewithGenerativeAI is a #WeeklySeries I am running on mentoring #TechStartups and #SmallBusinesses to leverage the power of AI to scale.
******************************************************
#reviewswithranjani #ScalewithAI Series
#Technology | #Books | #BeingBetter