The journey to adapt Sienna AI (2002â2025) for healthcare began in August 2022, after noticing the UK NHS (National Health Service) had a chaotic patient processing and staff allocation system that would benefit from an adaptation of the TBS-CC OKRs (Objectives and Key Results) System.
On the 23rd of March 2023, I noticed administration errors at my GP (Family Doctor) surgery, and proposedâat no cost, as part of my Innovate UK Grant applicationâto design and engineer a robust new administrative system. Alongside the note, "I hope in time it can connect to ChatGPT to answer 90% of all enquiries, first administrative, but in time medical as well." I called it "GP-AI".
On August 11th, 2023 I began what would end up as a 1000 page 240,000 word series called The Pregabalin Hangover. It began as a pain symptoms diary, but expanded into diagnostic conversations with GPT-4. And on Day 57 in Part 3 GP-AI asked the question "Can The GP-AI Outperform a Human GP?", having already concluded that GPT 4 outperforms psychiatry by +1000%.
In the spring of 2024, I compiled a 103-minute Podcast Episode, The GP-AI ðïžâïž. Taking on board the diagnostic power of AI but adding to it the Sienna AI franchising systems that could expand a bespoke concierge private-practise that would use GP-AI, across the nation.
2 â The GP-AI Project for Wes Streeting âïž
When Labour took power in July 2024 they paused the Innovate UK Smart Grant competition. So on October 8, 2024 I adapted the ð§6M Design from Business to Government and the NHS and stared the series ðŒ 2066 âïž The GP-AI Project for Wes Streeting ðïžð ð
2066a) âïž GP-AI Part 1. Revolutionizing Healthcare for an Aging Population âðªïž [09 Oct 2024] https://SiennaAI
.net/docs/GP-AI/2066a
2066b) âïž GP-AI Part 2. ALL-COMMs â David Farley â Microsoft â Open AI [09 Oct 2024] https://SiennaAI.net/docs/GP-AI/2066b
2066c) âïž GP-AI Part 3. Expanding NHS Efficiency with TBS-CC OKRs 4.7 â Creating a Competitive, Fun Environment in Healthcare ââš [09 Oct 2024] https://SiennaAI.net/docs/GP-AI/2066c
2066d) âïž GP-AI Part 4. OKRs 5.0 â Powering Government Objectives ââš [19 Oct 2024] https://SiennaAI.net/docs/GP-AI/2066d
While it may look simple in retrospect, GP-AI Part 3. Expanding NHS Efficiency with TBS-CC OKRs 4.7, and GP-AI Part 4. OKRs 5.0 â Powering Government Objectives were the evolution of the OKRs system we've been developing since 2017, work that could be seen at 2012: OKðRs 4.6 DevOps (+4.4, 4.7, & 5.0). including
2012t) The TBS-CC OKðR DevOps Approach to TDD Health Services in the UK âðªïž [14 Jun 2024] https://SiennaAI.net/docs/GP-AI/2012t
2012z5f) OKðRs 5f. The GP-AI Project and OKRs Reducing the NHS Waitlist from 7.6 to 2.3 Million by 2029 âðªïž [20 Oct 2024] https://SiennaAI.net/docs/GP-AI/2012z5f
there are 3 to 5 documents in 2075, in November 2025 2077 GP AI physio begins at almost the same time as The Good Doctor at 2078, with the good DR app Offshooting in GP-AI Psych, of which the video Terrifying lessons in Psychiatry brings us up to the 15th of december
2088 GPAI Gatekeeper begins on the 15th of december 2025
[Remember GP-AI medical record fraud is 2089 which deals with medical records]
2 â GP-AI Psych â Terrifying Lessons in Psychiatry âïžð§
With bias and criminal marketing eliminated GPT-4 radically outperformed psychiatry
3 â GMC ExWitâïžð§
The General Medical Council's Expert Witness Problem solved.
4 â OKRs 4.7âïžð
An adaptation of our OKR (Objectives and Key Results) System, to deploy across the UK NHS (National Health Service) incentivising performance, enabling real time performance analytics to governing bodies for every employee every day.
5 The Good Doctor Appâïžð
The Good Doctor App: Autistic Insight Meets A.I. Precision.
Specialist Knowledge Integration, Savant Reasoning, Medical Scans and Diagnostic Technology, Real-Time Monitoring, Expert Advice, and Precision Meets Guidance in the Operating Room, and went on to describe Complex Consultations, Unleashing the Potential of the Millennials, Training Simulations, and SURHs â 7 Star Superuniversity Resort Hospitals
6. GP-AI Gatekeeperâïžâïž
GP-AI-Gatekeeper "we're not in Kansas anymore dorothy" â theoretical ideas turned into engineering spec documentation for the Innovate UK Smart Grant competition. the 16 stage Design replaces the GP (Family Doctor's) receptionist with an AI following hierarchical decision trees and other methods to extract the perfect symptoms and history prompt. this is turned into a 400 word memo which is read back to the patient so avoiding voice to text errors and presented to the doctor and at the same time goes through the diagnostics and tells the patient exactly what's wrong with them And provides the GP with the specialist opinions they would usually send the patient to a hospital to procure. included the future concept that inspired Sienna Health EHR, where the patient would receive the result of the scan before they put their shoes on!
(Add the video thumbnails here!)
7. The £112 to £147 billion Economic Gain âïžð·
Welcome to GP-AI Gatekeeper 2025, a groundbreaking initiative by Sienna AI Ltd. to revolutionise healthcare, enhance public safety, and drive economic growth generating £112â£147 billion in GDP & savings from AI-driven healthcare.
7b. The GDS GOV.UK CMS Solution ðð»ð
The GDS GOV.UK CMS Solution ðð»ð was developed after a meeting with Helen Maguire, my local MP, who showed interest in GP-AI Gatekeeper. Since my background was in CMS design rather than NHS AI development, I thought it helpful to demonstrate how my CMS design could be applied to government projects. This journey revealed that the GDS (Government Digital Service) Gov.UK platform still operates on technology from 1998, seemingly for no reason other than a lack of understanding within the team. It felt fitting to show that what we achieved in 2002 surpasses the UK's current digital infrastructure. It also offered an opportunity to showcase what S-Web 6 VC AI CMS was, and its development journey since 2002.
7c. GP-AI Psych ð¬ð§ PART 7 â Suicide Blonde ðð±â
(I made a note that I should add this, but now I'm not so sure)
8. Expert-Diagnostics ð¬âïž
on the 26th of August 2025 a combination of GPT-5 Pro, Grok 4, Gemini 2.5 Proð, and Heidi Health AIâïžbegan the 2100i Expert Medical Opinions journey to creating an AI diagnostic that was as good if not better than the best doctor in the UK, in the field of Iatrogenic neurology, pain and anesthesiology.
By first creating the perfect prompt â symptoms in history using deep research then having the 4 models critique each other until they all concured on diagnosis and treatment.
leading to the 416-page
2100i13A5) ð¹âïžð§ Iatrogenic SFN â CSS â PEM â Fibromyalgia â A Multi-AI Clinical Research Dossier âðŠŒ [18 Nov 2025] https://siennaai.net/docs/GP-AI/2100i13A5
Consider the benefits of your having the very best specialist doctors opinion before you even speak to your GP (Family Doctor).
And consider the benefits for each specialist having the opinion of a peer specialists in other specialties who've had as much time as was needed to speak to the patient, before the consultation â that's what happened on the 28th of November 2025 with Nick Ray Ball and Dr Oliver Seyfried.
9. The-BrainSIM-Neurology-Projectð¬ð§
On the 30th of August 2025, whilst in the field of neurology the Expert-Diagnostics's research took a surprising turn whilst preparing to Present to Professor Mark Edwards at the Atkinson Morley Regional Neurosciences Centre.
it had been established that the conditions Sienna AI founder Nick Ray Ball suffers from have No cure And he will bed bound except for two minutes a day for the rest of his life, of course Nick Grey Ball wondered about creating a cure, the problem was caused by a drug reaction so couldn't another drug reaction reverse the problem.
The consideration was because A is are trained in neural networks that to a degree mimic the human brain could we make a copy of the brain and use that to test treatments? that wasn't possible but one could create a simulation of the disease state of the brain given about 1% of a major AI Labs infrastructure.
the result was that in theory neurology could move from a hundred therapeutic attempts a year to a billion â without putting a single extra human at risk?
On the 20th of september with CA-125 scan result test result of 93 Monica Ball, Nick's Mother, faced an uncertain future and cancer. But there were mistakes made, there was no reminders to take tests, and things that would take an AI less than a minute were taking weeks.
stage 16 of GP-AI Gatekeeper, presented the future concept of receiving the results of the scan before you've put your shoes on. So how could we automate the cancer track replacing human bottlenecks with instant AI analysis and decisions alongside the diagnostics and the gpai with the gpai taking a greater role than the receptionist rather becoming an individual medical record database of each individual patient That various A could access.
remembering after a decade of making business administration systems in the real world in 2011 Nick Ball would spend fourteen years working on the economics of AGI. a network of referral businesses that would using various macroeconomic technology designs generate more than half of the global GDP.
so simplifying that greatly to just a system to run the patient journey through hospitals and the healthcare system was hugely simpler.
and it was discovered that to get that cancer track down from three months to three days you would need to create a new EHR electronic hospital record system.
without cooperation from Epic and Oracle, this will be logistically very difficult, but given the savings and increased GDP for the USA at approximately 3.9 trillion a year, the Trump administration and the Democratic Party could easily afford the 10 billion a year the current companies make to incentivise them to recreate the system in a way similar to Apple's iOS, micro services with apps that can plug straight in.
if this could be accomplished that three month to three day journey is very real.
âTaking the savings to Medicaid and Medicare and by assisting 4% of the US population to work who currently cannot, the total GDP boost and savings to the USA per year is conservatively $3-3.6T.â
11. GP-AI 2.0âïžð
On the 2nd of October 2025, I had a follow up meeting with Dr Deji Akiboye, the conversation went well and drifted to the GP-AI, I asked Deji if he like to see the videos and he said he would. Deji then explain something he'd been trying to work out using algorithms Which was the problem that specialist doctors in hospitals find it almost impossible to access a patient's medical history even the previous special.
Three days later I created 2110e1A) âïžSienna Health EHR, a Hawking winning document. a Hawking (âðŠŒ) is the highest scoring emoji in the OKR System worth 2,500 points.
In between the 8th and the 11th of October I created:
2110d1B) âïžSienna Health and the GP-AI Project for Dr Deji Akiboye ââš [8 Oct 2025] https://SiennaAI.net/docs/GP-AI/2110d1B
2110d3A) ðïžâïžSienna Health â How the GP-AI Would Coordinate Specialist Care for Dr Deji Akiboye ââš [11 Oct 2025] https://SiennaAI.net/docs/GP-AI/2110d3A
Both scoring the plank star ââš, some weeks I'll not see a Planck star or a hawking, a Hawking and two Planck stars in 9 days is a great haul. and for those who are interested, yes, gamifying the OKRs leads to an awful lot more work.
In the end, the solution to the specialist doctors not accessing patients history problem was simple to fix, if you already have the design for the GP-AI. In Sienna Health USA we'd advanced the GP-AI's role as the receptionist who gives diagnosis, and GPs specialist opinions. To the individual patients AI, no longer was it the GPS AI but rather the patient's AI GP (family doctor), connecting to the EHR. Theoretically reducing the cancer track from three months to three days.
To solve the problem posed by Dr Deji Akiboye, GP-AI needed to become, the database that contains the patient's medical history, which involves AI scanning each patients 250-5000 pages of medical record across all record keepers, sorting it by date, doctor & condition, with all keywords tagged for searches and important aspects summarised. which isn't that hard, once you have access. This would allow any specialist to ask Sienna "please give me all previous urological results, test the pharmaceuticals I want to prescribe against all currently used pharmaceuticals for conflicts, and give me any other information that is relevant." And voila! Sienna (GP-AI Gatekeeper) replies.
Future investigations unearthed a big contrast between the UK and the USA, where the patient's medical record is held by the NHS GP in the UK, in the USA it is held across the hospitals. making the GP-AI even more invaluable to the USA than the UK.
Back in the UK GP-AI would become the record keeper at the NHS GP Surgery, instantly accessible by all specialists across all hospitals and healthcare institutions. And the first thing we do as soon as it was collated would be to read the key things back to the patient for the patient to correct the errors, which if my case is anything to go by will be many, this single step will save lives and possibly yours!
So the GP-AI evolves from an AI into a database accessible by AI, formatted for maximum efficiency, searchable by keywords and general requests. And in the USA immediately before any ideas of creating SiennaHealth EHR, the GP-AI can plug straight in to Epic, Oracle and other EHRs, as was explored in
2100i17A2D4b) âïžâïžð§ Beyond Part 3 â From Patient Chart to Patient Brain â Why GP-AI Must Sit Above Epic & Oracle â Not Inside Them [29 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A2D4b
12. BrainSim2-Drug-Discoveryðð¬
On the 23rd of November 2025...
2100i12E15) âïžâïžð§ The BrainSIM Neurology Project Book 2 â The Big Question About Drug Creation âðªïž [23 Nov 2025] https://siennaai.net/docs/GP-AI/2100i12E15
2100i12E16) âïžâïžð§ BrainSIM 2 â The AGI-Assisted Neurological Drug-Discovery Engine â Summary ââš [23 Nov 2025] https://siennaai.net/docs/GP-AI/2100i12E16
...
2100i13B3) âïžâïžð§ The BrainSIM Neurology Project â 10Mà Variant Acceleration + AGI-Assisted Drug-Discovery âðŠŒ [24 Nov 2025] https://siennaai.net/docs/GP-AI/2100i13B3
13. The N-of-1 Drug Variant Trial for PAIN Reports (IASP) Journal ð§âð¬ð
On the ?? of November 2025...
2100i17A2) âïžâïžð§ Nick Ball â Iatrogenic SFN â CSS, PEM & FM â Chronic Pain Symptoms & AI Clinical Overview for Dr Seyfried at SGH âð [8 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A2
2100i17A2B) âïžâïžð§ Part 1 â The Importance of Documenting Nick Ballâs Functional Limits for Dr Oliver Seyfried (SGH) (Severe SFN â CSS â PEM â FM) âð [8 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A2B
2100i17A2D) âïžâïžð§ Part 3 â Treatment Research â Diazepam â Duloxetine â LDN for Dr Seyfried (SGH) (Iatrogenic SFN â CSS â PEM â FM) âð [8 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A2D
2100i17A3) âïžâïžð§ SFN â CSS, PEM & FM â Patient-Led Tri-Modulation Trial â Diazepam ⢠Duloxetine ⢠LDN for PAIN Reports (IASP) Journal [10 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A3
2100i17A5B) âïžâïžð§ State-dependent SFN Symptoms â Link Between Consciousness and Freezing-Feet Onset â Diazepam. (Questions 1 to 4) âðªïž [24 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A5B
2100i17A6) âïžâïžð§ Generative AI pre-trial protocol â N-of-1 Diazepam â Duloxetine â LDN Trial for Iatrogenic SFN â CSS â PEM â FM âðŠŒ [24 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A6
2100i17A7) âïžâïžð§ Iatrogenic Central Neuropathic Tinnitus â thalamocortical gain disorder secondary to ototoxicity âðªïž [25 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A7
14. SiennaHealth.AI
On the 28th of November 2025...
2100i17A2D3) âïžâïžð§ Beyond Part 3 â From Treatments Research with Dr Seyfried to SiennaHealth.AI EHR UK UCLH SGH Discussion [28 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A2D3
2100i17A2D3b) âïžâïžð§ Beyond Part 3 â From Patient Chart to Patient Brain â Why GP-AI Must Sit Above Epic & Oracle â Not Inside Them âðªïž [29 Nov 2025] https://siennaai.net/docs/GP-AI/2100i17A2D4b
2110i1A) ð°ïžð§ âïž SiennaHealth.AI â Realising the Architecture of an Entirely New Medical Era. [29 Nov 2025] https://siennaai.net/docs/GP-AI/2110i1A
