Mohit kumar

Drink from the well, replenish the well.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft I just love it when Shankaran said team is living, breathing, eating, sleeping chandrayaan for last 4 years to complete this. This is the kind of effort which is key to every project.

[Mohit kumar @MohitKumar6727]

@raydipankar92 @shubhsinghnft https://t.co/ErVOaOTDUq

https://t.co/YeIQpy29FP

check these

[Mohit kumar @MohitKumar6727]

@harrywalia019 @shubhsinghnft @kunikhanna @ApeEvent1104 Yeah, looks like there is no working area for a medium size gathering for a day on a weekday.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft @harrywalia019 @ScribeHow Sure

[Mohit kumar @MohitKumar6727]

@shubhsinghnft @harrywalia019 @kunikhanna @ApeEvent1104 @famboytweet @JWMarriottChd @Hyatt Yeah, going out tomorrow. Tomorrow is site visit day 🤘

[Mohit kumar @MohitKumar6727]

@shubhsinghnft @harrywalia019 @ScribeHow It's not completely autonomous. It required human touch to add some context to the documentation but all time that is consumed in creating base structure was saved.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft Pretty cool

[Mohit kumar @MohitKumar6727]

@shubhsinghnft @harrywalia019 @ScribeHow Check your email. The documentation you needed for epharmacy, due to time crunch, we found this tool and created all that within minutes.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft Yes

[Mohit kumar @MohitKumar6727]

@shubhsinghnft True, I agree 💯

[Mohit kumar @MohitKumar6727]

@kunikhanna @shubhsinghnft Exactly, that's what I am talking about.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft @kunikhanna I agree. And not just that, where there is digitisation to some extent, the systems are independent from each other. Data is sitting in silos. Interoperability is not there even in developed countries.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft Actually based on my experience, Amazon's textract is better than vision. It gives more control to you on different format of extraction and what to do with the extracted information.
Yeah, patient files which are low quality or not maintained, we are not there yet for OCR.

[Mohit kumar @MohitKumar6727]

@kunikhanna @shubhsinghnft If it's digitised, there can be measures in place and control can be given to user to whom they want to share it with.

[Mohit kumar @MohitKumar6727]

@shubhsinghnft It can read but out of context. I think the main issue we will have is making the sense of the scanned data.

Also handwriting is not the only issue, the state of the physical files they are in, old, bad paper and so, also posses a challenge

[Mohit kumar @MohitKumar6727]

@shubhsinghnft We have tried Tesseract, Amazon's Textract, google vision AI. All of them gave good results. It can handle hand written notes very well.
@shubhsinghnft

[Mohit kumar @MohitKumar6727]

@shubhsinghnft To harness AI's potential, we must prioritize efforts to digitize medical records and create interoperable systems. I think with the E-health TT platform, that's a step in the right direction.