MyBotGP Docman Filing Webinar Recap: Key Takeaways and Q&As
On the 28th of January 2025, we hosted a webinar on our product—MyBotGP EMIS Docman Filing. This innovative solution is set to be a game-changer for NHS Primary Care, offering an automated approach to managing document filing. During the webinar, we answered a variety of key questions from attendees, which we’ve summarised below to give you a clearer understanding of how this product works.
Q1. Do you help with DCB 0160 paperwork?
A) Yes, we can support with DCB 0160.
Q2. Does it recognise that it is a letter about safeguarding and make it non-visible online?
A) It’s how you classify the letter; it has the ability to select or deselect that type of letter to make it visible only.
Q3. Are you saying you have to train it on every ‘type’ of letter—how long does that take? Can you just tell it to send every letter to one particular inbox without having to train?
A) Yes to both.
Q4. Do you have real practice/PCN accuracy data for the automation?
A) Yes, we can share the details. We have our DCB 0129 certification. In case there is an issue and it’s misfiled, there are also protocols around for making the issue resolved easily. We can make enhancements online.
Q5. Is it using image-to-text to ‘read’ the letter and extract the right info?
A) Yes, lots of different technologies are used. A plethora of technologies are put together for its working.
Q6. How does it cope with multiple options picked up by IntelliSense?
A) You supersede it and tell it what is correct.
Q7. What is the cost of this project? What is the accuracy level in %? What kind of support is available post go-live?
A) Cost is £749 + 30 pence per patient and hosting cost per patient. We run training after installation and have weekly drop-in webinars. In those webinars, we have trainings and tips; we also have videos and trainings online.
Q8. Will it pick up letters we have manually filed into Docman that have not come in via email?
A) Yes, we need to scan and save it in the folder. The only manual task.
Q9. How are exceptions handled? How many documents/pages can it process in an hour? Does it use any API to interface into the clinical system (EMIS)?
A) It takes around 5 to 10 seconds per letter max; it uses EMIS the same way a person does.
Q10. I am assuming that for any new letters, new codes, or new hospital names, we need to train the bot or update. Please correct me if I am wrong.
A) Yes, it takes about one to two minutes; the second time they will know what to do.
Q11. Are there any videos available so we can see how this bot works?
A) Yes, we do have videos available. You can check them out here.
Q12. Letters with several different dates on them—how will it know which is correct?
A) We show it to recognise the letter, like where the date is placed, with what the keywords are; the second time it will know the placement and track the date.
Q13. What if the letter has third-party names that the bot does not recognise?
A) Yes, we can train it and it will recognise a third party or private consultancies.
Q14. Can you file the document to more than one patient record, e.g., safeguarding for multiple children? And can it link with redaction software?
A) Yes, we can file; in terms of redaction, no one looks at the patient’s data.
Q15. If there’s an existing template already in place, can it be set up to recognise key words and change procedures in specific circumstances? (e.g., safeguarding concerns raised)
A) You would train the letters to make them look for those keywords; it’s all about the training given to the software.
Q16. We send our letters to the usual GP, so 10 copies of the same letter type may need a review created for different GPs. I presume that this isn’t possible with your system?
A) Yes, you can train and put it to different GPs. If a particular GP is looking at a particular thing, it can work accordingly in sending to a specific person.
Q17. Is this essentially in place of a person that is scanning and then distributing the documents and not workflow teams that are reading the documents and coding and actioning, etc.?
A) Yes, it replaces the manual scanning and distribution process.
Q18. Do you have a DTAC and DPIA template in place?
A) Yes, we do.
Q19. If you have two letters from the same department, one with actions and one without, is it able to create a review for the letter with actions and file and not ask the one without actions?
A) If you can recognise it clearly from the letter, then no issue, or else it’s all about the training given.
Q20. Can you please explain how MyBotGP deals with letters we get in the post?
A) The system would scan and email them anyway; only handwritten would be handled by the team; the rest would be handled by the system.
Q21. What about handwritten letters—can it read them?
A) Yes, if scanned and uploaded.
Q22. Can this be scheduled at any time?
A) It is scheduled to run three times a day—once in the morning, once in the afternoon, and once in the early evening.
Q23. Does it recognise if multiple patients are on one document, i.e., Mom and children?
A) It would pick the letter type and send it to the appropriate folder.
Q24. Will you have the same integration in NCL ICT as you do in HBL ICT?
A) We can access everyone nationally, across the country.
Q25. Is there an audit trail in case the letter is filed into the wrong patient’s record?
A) Yes, there is; it is redundant now.
If you have any questions or want to book a demo, please submit the form here: https://jifjaff.co.uk/mybotgp-docman-filing/ or email us at enquiries@jifjaff.co.uk.