Iโm giving away a new iPhone 18 Pro Max ๐๐ฅ
In exactly 24 hours, Iโll pick 8 random people from my replies to win.
RULES: like and repost this post, comment โReadyโ and follow me.
WORLDWIDE SHIPPING ๐
Thanks and good luck โค๏ธ
Iโm giving away a new iPhone 18 Pro Max ๐๐ฅ
In exactly 24 hours, Iโll pick 8 random people from my replies to win.
RULES: like this post, respot, comment โReadyโ and follow me.
WORLDWIDE SHIPPING ๐
Thanks and good luck โค๏ธ
@DonaldLamp84do This is exactly the failure I'm scared of, cuz a fluent answer on a missing clause sounds like an approval. Planning 3 states per answer: quoted (with clause), inferred (flagged), not found (escalate to insurer). Will test it on questions with no answer in the policy.
guide me...
Day 01 of building an AI voice agent ๐๏ธ
Nobody reads insurance fine print until a claim gets rejected.
So I'm building an agent that reads the brochure and explains your coverage, out loud.
#BuildInPublic#AI#VoiceAI#Insurtech#Golang
Hey seniors, I am in my 6th sem, and I have heard that there will be no summer vacay for the people in the 3'2 and instead there will be placement training including Saturdays and Sundays. Could you tell to what extent this is true? And if it is true how did you guys survive this
@kaggle My U-Net model is officially learning to spot forgeries in biomedical images for the @Kaggle comp.
Left: The original (forged) image Right: My model's predicted mask
The validation loss is dropping. So satisfying to see it learn.
#AI#PyTorch#DeepLearning#ComputerVision
@kaggle Had to build an integrity check that filtered out 882 corrupt/mismatched image-mask pairs. ๐คฏ Now training the U-Net on 1,869 clean samples.
Never underestimate data cleaning.
#Kaggle#ML#DataScience#Python