Great blog article Ashu. Why not create an agent (per your signals above) but add a storage layer that persists all data points recorded during that agent's execution. So initially you start with a non-context driven agent, but allow its actions to be over-ridden by humans. Record all data points in either case. Over some time, you can run ML algorithms to decide colinear factors, algorithms weights and so on. Then the agent can start recommending exceptions.
In that manner, slowly build up a context graph. Why do we need a new company for this?
Hi @Ashu Garg your context graph article directly plagiarizes from my process knowledge management series.I have reported this to @Substack Team as it violates the policies of the platform
You also plagiarized from the work embedded within my citations.
I expect we can resolve this with attribution and citations to my work and the work of others, cited in my work?
Great article, Ashu!
Great blog article Ashu. Why not create an agent (per your signals above) but add a storage layer that persists all data points recorded during that agent's execution. So initially you start with a non-context driven agent, but allow its actions to be over-ridden by humans. Record all data points in either case. Over some time, you can run ML algorithms to decide colinear factors, algorithms weights and so on. Then the agent can start recommending exceptions.
In that manner, slowly build up a context graph. Why do we need a new company for this?
Hi @Ashu Garg your context graph article directly plagiarizes from my process knowledge management series.I have reported this to @Substack Team as it violates the policies of the platform
You also plagiarized from the work embedded within my citations.
I expect we can resolve this with attribution and citations to my work and the work of others, cited in my work?
This is theft.
we have something like this working at salley.co.
excited to go live at a major enterprise soon :)
Fantastic