For pharmacists
Spend your time on decisions, not on finding them.
A practice pharmacist’s day is full of reading letters for medication changes, reconciling them against the record, preparing requests and chasing monitoring. Reedley is being built by a pharmacist to carry the reading, checking and chasing — so your time goes on the clinical decisions.
The demo opens the same synthetic discharge summary.
In practice
One reconciliation, done the Reedley way.
A synthetic discharge summary, from the worklist to a decision you make. Intended behaviour of a product in development; nothing here changes a record.
01 Worklist
- Graham HOLT NHS 999 000 0417Reconcile the medication record with the discharge summaryMedication reconciliation · prescriber review requiredR. SHAH, clinical pharmacist · due within 2 working daysProposed · awaiting review
- Patricia LOWE NHS 999 000 0387Methotrexate monitoring · FBC, LFT, U&E dueMonitoring due · draft rule, unsignedAdmin team · due 13 OctAssigned · due 13 Oct
- Funmilayo ADEYEMI NHS 999 000 0069Review the ramipril dose change requested by cardiologyMedication change · prescriber decision requiredK. AHMED, clinical pharmacist · due todayReviewed · owner to be set
Three synthetic items, each at its own moment. The worked example is the discharge summary for Graham HOLT: Discharge summary · Northfield General Hospital (fictional) · received 3 Aug 2026 16:40.
02 Review the evidence
Shown, not resolved. Nothing in the record has changed.
Proposed work · Proposed by Reedley · 4 Aug 2026 09:14
Reconcile the medication record with the discharge summary
- Kind
- Medication reconciliation
- Owner
- R. SHAH, clinical pharmacist
- Review level
- Prescriber review required · floor: draft, awaiting CSO sign-off
- Due
- Within 2 working days of receipt (by 5 Aug)
- Verification
- How it would be verified: the medication record re-read with approved access, once (Simulated on this site)
Proposed · awaiting review
03 Decide
Task lane · Medication reconciliation
Reconcile the medication record with the discharge summary
Owner R. SHAH, clinical pharmacist · Prescriber review required · floor: draft, awaiting CSO sign-off
With JavaScript the review controls can be tried here. Accept records “Reviewed by R. SHAH, clinical pharmacist · 4 Aug 2026 09:32” and the item moves to assigned, never further; Edit changes the owner or the due date; Reject needs a structured reason.
Decision lane · not a task
¶4 “Please arrange urea and electrolytes within 7 days of discharge and review the result before considering restarting ramipril.”
Routed to Dr S. PATEL, GP · a GP decision · no medication task
Reedley does not decide whether to restart a medicine. The GP’s review and decision are recorded as an event; what was decided is not shown here.
04 Status, capped at assigned
Identified · proposed, awaiting review
Your decision moves this item to assigned at most. What happens next is outside Reedley:
- You, in the clinical systemYou make the changes in EMIS Web or SystmOne. Reedley does not change the medication record.
- Reedley · design intentRe-reads the medication record with approved access (none sought yet; Simulated on this site) and shows what it finds.
- Verified · 5 Aug 2026 15:10 (Simulated)Medication record now matches the discharge regimen. Limit: issued is not taken.
Reedley does not change the medication record. A prescriber makes the changes in the clinical system. Reedley does not decide whether to restart a medicine.
After assigned: one monitoring item, from requested to reviewed.
The renal function check that follows a dose change, as the six states would show it. The owner’s request leaves it at actioned; the received result verifies that the test happened; it is designed to close only when the clinician’s review is recorded (the development build closes it on receipt).
Illustrative · the intended lifecycle
Identified
Reedley has found an item of work and attached its source evidence: a cited passage, a rule or a received result.
Entered by System
4 Oct 2026 · 10:05 Identified from ¶3 of the clinic letter: “recheck renal function in 1–2 weeks”.
Reviewed
A named person has examined the evidence and accepted the item: accept, edit or reject with a reason.
Entered by Clinician or delegated reviewer
4 Oct 2026 Evidence checked against the letter and the record; task accepted.
Assigned
A named owner and a due date exist.
Entered by Reviewer or team lead
4 Oct 2026 Owner: Admin team · window 11–18 Oct, 1–2 weeks after the dose change recorded 4 Oct.
Actionedreported, not proven
The owner reports the action is done. Reported, not proven: the item waits here while evidence is awaited.
Entered by Owner (a person)
5 Oct 2026 U&E requested 5 Oct.
Verified
Evidence from the relevant source of truth shows the action itself happened. The evidence line says exactly what it proves.
Entered by System, from source evidence (Simulated). A person’s report moves an item to actioned; verification comes from the record where Reedley can read it; where it cannot, a named person’s structured statement is accepted today, and that rule is under review.
14 Oct 2026 · 09:12 Pathology result received into the record.
Closed
The obligation ended, after verification. Three outcomes: completed · unable to complete · not required.
Entered by System (completed); a named person with a reason for the other two outcomes
14 Oct 2026 · 16:10 Clinician review recorded in the clinical record.
Funmilayo ADEYEMI NHS 999 000 0069
Renal function (U&E) 1–2 weeks after the dose change
Investigation · administrative
Show all eventsHide the event list(6)
- 4 Oct 2026 · 10:05ReedleyIdentified from ¶3 of the clinic letter: “recheck renal function in 1–2 weeks”.Source passage cited (¶3). Depends on the ramipril dose change being recorded.
- 4 Oct 2026K. AHMED, clinical pharmacistEvidence checked against the letter and the record; task accepted.
- 4 Oct 2026K. AHMED, clinical pharmacistOwner: Admin team · window 11–18 Oct, 1–2 weeks after the dose change recorded 4 Oct.
- 5 Oct 2026Admin teamU&E requested 5 Oct.Reported by the owner · not yet proven.
- 14 Oct 2026 · 09:12Reedley · from pathology as received in the record (Simulated)Pathology result received into the record.Result received 14 Oct 09:12 · test completed · not yet reviewed (Simulated).A received result is not a reviewed result. Designed to stay open until a clinician’s review is recorded (the development build closes it on receipt).
- 14 Oct 2026 · 16:10Reedley · from the clinical record (Simulated)Clinician review recorded in the clinical record.Reviewed by K. AHMED 14 Oct 16:10 (Simulated).Shows that a review was recorded, not what it concluded. Reedley does not interpret results.
Awaiting a named person’s review.
Owner and due date to be set.
The owner arranges the blood test and records what was done.
Waits here until a result is received from pathology.
A separate review item goes to K. AHMED, due today.
Closed · completed after the review was recorded.
Illustrative concept · not clinical rules · source checks are simulated; automatic checking needs approved access to each source.
No practice uses Reedley yet · Tested on synthetic data only · No EMIS Web or SystmOne integration claimed
Designed for you
Built around how the work is really done.
Every change cited
Select a finding and the exact passage opens beside it. Nothing is paraphrased without its source.
Unclear wording flagged
“Consider stopping…” becomes a clinical review, and Reedley can draft a clarification to the specialist for a clinician to send.
Done means done
Designed so that work closes on evidence, not on a tick. Checking the record automatically will need approved access.
Your decisions recorded
Who decided, when, and on what evidence — kept in an append-only history.
Routing by capability
Prescriber-level work reaches prescribers; administrative steps reach the team that books and files.
Rules you can see
Rules are designed to be versioned and owned by a named clinician. Today every rule is an unsigned draft with no named owner.
Honest status
Where the relevant capabilities stand.
As at 14 November 2026. Nothing here is in use with real patients.
| Capability | Status | Before real use |
|---|---|---|
| Instruction comparison (a classifier in the development build’s domain code) | Built · synthetic data | No screen of its own; dm+d-backed terminology and real-letter evaluation outstanding |
| Medication reconciliation workspace (side-by-side review, per-line decisions, action list, handover) | Planned | Not built: shown as an illustrative demonstration; depends on a decision about how a clinician’s account of an action is recorded |
| Restart context and instruction conflicts | Built · synthetic data | Synthetic records only · instruction conflicts: Synthetic prototype, collated for a prescriber’s review |
| Prescription request preparation | Needs clinical sign-off | Synthetic prototype · sorts for a prescriber’s review · sorting rules need clinical sign-off; no route to the real request queue |
| Monitoring rules | Needs clinical sign-off | Draft rule table with no clinical owner yet |
| My Day: your worklist, with the reason for every item | Built · synthetic data | Core platform · intended launch scope · in-memory in the browser; no clinician has used it |
| Patient Context and What Changed | Built · synthetic data | Core platform · intended launch scope · prioritisation weights hand-set; no real-record evaluation |
| Ask Reedley | Built · synthetic data | Core platform · intended launch scope · answers from the record with evidence on every line; no language model is consulted in the current build; accuracy not measured |
| Writing to the medication record | Withheld by design · not built | Not built. Any future write would need supplier approval, CSO sign-off and a device determination |
The platform pages use three public states for the same facts: “Built · synthetic data” and “Needs clinical sign-off” are in the development build (synthetic data; draft rules unsigned), “In development” is partly built, and every other status here is not built. An illustrative demonstration can show parts that are not built.
Where you would work in the workstation: My Day · Medication Reconciliation · Patient Context · Ask Reedley
Not sure where to start? Take the five-question assessment
Will Reedley change medicines in the record?
No. It shows what the letter asks to change and is designed to check afterwards that the change was made. The prescriber makes the change in the clinical system.
Does it work with EMIS Web or SystmOne?
It is designed around workflows in both, but no integration exists or is claimed yet. See the integration page for the honest status.
Who built it?
Reedley is being developed by a UK pharmacist with years of experience processing prescriptions and correspondence in general practice.
Pilot interest
Help shape Reedley before it is finished.
We are looking for a small number of GP practices and primary care networks to work with us: first on synthetic walkthroughs, then — only with the right approvals, safeguards and agreements — on carefully governed evaluation.
- 01Register interestA short form. We reply personally to arrange a conversation. No commitment. Or request a demonstration
- 02Synthetic walkthroughYour team uses Reedley on synthetic patients and tells us what is wrong with it.
- 03Structured feedbackTimed tasks and short interviews, reported honestly, including what did not work.
- 04Governed evaluationOnly with a CSO-approved safety case (DCB0129), the practice’s DCB0160, an MHRA device determination, DSPT, a DPIA, a data processing agreement and supplier approvals in place.