“This is NOT a list of hackathon ideas. It’s a list of items where companies are currently paying $$$.”

Lucy Park, Co-founder & CSO of Upstage

OpenAI gets an enormous amount of capital, press and buzz for being first to market with the LLM. The question of how they will earn revenue anywhere in relation to their market cap is TBD.



Lucy Park, Co-founder & CSO of Upstage a Korean AI firm, spoke yesterday about their incredibly pragmatic approach to building AI solutions companies had a need for and WILLINGNESS to pay them to solve.

They started with computer vision projects to take documents with unstructured data, translate it into formats that can be chunked and embedded in a database and then searched.

Upstage initially focused on the mundane, common and highly repetitive use cases that make for dead end human jobs with low job satisfaction and high error rates:

Process automation
– document classification
– Information Extraction

Productivity enhancement
– summarization
– writing assistance

Knowledge search & discovery
– enterprise search
– business analysis
– customer support

I had been talking to another woman at the conference who was transitioning from software engineering to AI and wanted to find learning methods that cut through the hype to the real methods and applications.

Moments before Lucy’s talk I had made the hyperbolic statement “it’s all hype” and tinkering and rapid prototypes that will bring massive pain in production.

I’ve seen some use cases in my consulting work like what Upstage is providing which are proven and hugely valuable, but they don’t get the air play they deserve.

Vin Vashishta you would have been jumping out of your seat, it was so satisfying to see the fundamentals get their moment on stage.

A friendly reminder that it is an option to choose your AI & ML projects based on a business case of expected feasibility and ROI not FOMO!

Leave a comment