Sri Venkateshwara Multi-Speciality Hospital · Kukatpally, Hyderabad · 60 beds
KR

Good morning, Dr. Ramesh

Saturday, 18 Jul 2026 · General Medicine · Room 3 · OPD session 09:00 AM – 01:00 PM

Now: 10:47 AM Running 12 min behind

Today's queue

Token, patient, status & complaint for this session
next 7 of 26
T-12In room
Lakshmi Devi 58F T2DM follow-up
T-10Done
Venkatesh Naidu 44M Hypertension review
T-14Waiting
Mohd. Irfan 32M Fever 3 days
T-15Waiting
N. Mahesh 39M Cough & cold
T-17Waiting
D. Prakash 51M Knee pain
T-21Booked
G. Ramulu 63M Diabetic review
T-24Booked
S. Yadamma 70F BP check
Consults done today
14/26
54% through the list
Avg wait time
14 min
▼ 3 min vs last Saturday
Avg consult length
8 min
target 10 min
Running behind by
12 min
since 10:15 AM

Consults per hour

Today's session · 09:00 AM – 01:00 PM
Consults completed
3 4 3 4 so far 9–10 AM 10–11 AM 11–12 PM 12–1 PM

Wait time this week

Minutes, average per session
Mon Tue Wed Thu Fri Sat
17 min avg on Wed (clinic double-booked) · today trending better at 14 min
✦ Before you walk in — Lakshmi Devi AI briefing

✦ Reference & consistency check

Her recorded values against published ranges, and against her own history. Alerts, never instructions — nothing here is a diagnosis and nothing changes a prescription.
AMBER · reference

Live consultation — Lakshmi Devi

Token T-12 · Room 3 · in room since 10:41 AM
Recording · 02:41 Telugu + English
✦Tap ■ Stop & draft above — the AI Scribe will transcribe the Telugu + English consult, then write the SOAP note and prescription here for you to review.

AI Health Memory — Lakshmi Devi

Ask anything about this patient's history · answers cite the source visit
✦ AI memory
✦ 4 years of Lakshmi Devi's record indexed — 6 OPD visits, 1 admission, 9 lab reports, pharmacy refills. Tap a chip or type your own question.

New order — Lakshmi Devi

Token T-12 · sends directly to Lab & Pharmacy queues

Lab

Imaging & screening

Admission

Admit to ward

Saved templates

Disease-specific order + Rx bundles

T2DM follow-up

Labs: HbA1c, lipid profile, KFT, urine ACR
Rx lines: Metformin, Telmisartan — dose fields blank
Used 142 times
General Medicine

Hypertension review

Labs: KFT, lipid profile
Rx: Telmisartan 40mg
Used 98 times
General Medicine

Viral fever

Labs: CBC, malaria/dengue panel
Rx: Paracetamol, Amoxicillin-Clav 625mg
Used 211 times
General Medicine

✦ Prescription assistant — Lakshmi Devi

Formulary, generics, her allergies, and Telugu printing. The prescribing is yours.
GREEN · lookup & printing AMBER · allergy & label alerts

Formulary & generic equivalents · what 30 days costs her

30 days costs herDrugBrand ₹/unitGeneric ₹/unitStock here
₹161
120 tabs incl. 12% GST
Tab. Metformin 500mg in formulary — ₹1.20 420 tabs
₹151
30 tabs incl. 12% GST
Tab. Telmisartan 40mg in formulary — ₹4.50 38 tabs · below reorder
₹319
30 caps of the brand she is dispensed
Cap. Pregabalin 75mg generic available ₹9.50 ₹3.20 both stocked
₹631
her total, all out of pocket
Star Health policy on file — used for her Aug 2023 admission, carries no OPD benefit. Not enrolled in Aarogyasri or PM-JAY. Switching the Pregabalin line to the generic would move ₹631 → ₹420. Which one she gets is your prescription, not ours.

Dose — the label's numbers, and an empty field

Formulary defaults for this molecule: 500mg / 850mg / 1000mg, usual maximum 2000mg/day. Renal bands in the licensed label: eGFR 45–59, 30–44, below 30. Her recorded eGFR is 44. HOS will tell you which band that is. It will not put a number in the box — a dose chosen for a named patient is a RED-tier action under the CDSCO SaMD framework, and HOS is not licensed for it.

Vernacular printing

Print dosage instructions in Telugu · you already do this on 92% of prescriptions
మెట్‌ఫార్మిన్ — రోజుకు రెండు సార్లు, భోజనం తర్వాత. మాత్ర ఎప్పుడూ ఆపకండి, ముందు డాక్టర్‌ని అడగండి.
Metformin — twice a day, after food. Never stop the tablet without asking the doctor first.
టెల్మిసార్టన్ — ఉదయం ఒకసారి, ఖాళీ కడుపుతో. మందు అయిపోయే ముందే మళ్లీ తీసుకోండి.
Telmisartan — once in the morning on an empty stomach. Collect the refill before the strip runs out.
✦Sources: this hospital's formulary and live batch stock, the licensed prescribing information for each molecule, her allergy record, and the ABHA-linked Apollo Rx. Time honesty: across your last 40 prescriptions the assistant saved a median of 1 min 5 s each — mostly the formulary and generic lookups you used to do in another tab. It saves nothing on the decision, because it does not make one.

✦ Patterns across your consultations

The Scribe reads every approved transcript — across visits it starts seeing what one visit can't show. Private to you; insights are suggestions, never report cards.

Notes drafted today

14 of your 14 completed consultations have a draft · 4 still need your signature
GREEN · transcribe & format
2 min 10 s
Median saved per note — measured on your own 412 notes as draft-shown to your-signature, against your typed notes before the Scribe
30 min
Back to you today — 14 notes × 2 min 10 s. Volume is what compounds, not the percentage
0
Notes still open when you left, last week. In January it was 17 — the change your family notices
16.0 min
Per 8 scheduled patient hours — the published independent figure, shown so you can check ours against it
JAMA, 1 Apr 2026 · 8,581 clinicians, 5 systems
✦What we will not tell you. Vendors in this market advertise 1–2 hours a day saved. That figure is marketing; you would catch it by Friday, and then you would stop believing the rest of the screen. Ours is a measured median per note plus your own volume, and the published number beside it. If the two ever disagree badly, trust the published one.

Awaiting your signature

Nothing here is filed, sent or billed until you sign it
4
PatientDraftSeen forWhat you usually change
Venkatesh Naidu
T-10 · 44M
S · O · A ready Hypertension review Nothing in the last 8 — your BP phrasing is already learned
K. Sarojini
T-03 · 61F
S · O · A ready Thyroid follow-up You add the next TSH date — the draft now does it
Md. Ghouse
T-07 · 29M
S · O · A ready Fever, day 4 You name the day of illness explicitly — learned, 11 times
Lakshmi Devi
T-12 · 58F · in room
S · O · A ready · P empty T2DM follow-up Cannot be signed — no plan has been spoken yet
✦Tap Open the note on Venkatesh Naidu above — the drafted SOAP note appears here for you to read and sign. It is drafted from the recorded encounter and from values already in his chart; nothing in it is generated from anywhere else.

✦ ICD-10 codes proposed from your note — Lakshmi Devi

Each code names the line of the note it came from. You confirm; HOS files nothing on its own.
GREEN · coding Draft — awaiting approval
Code & your callTermWhere in the note it came fromConfidence
E11.40
T2DM with diabetic neuropathy, unspecified A: "diabetic neuropathy, symptomatic" · problem list since Nov 2025 · Pregabalin on the current list High
E11.65
T2DM with hyperglycaemia O: RBS 214 mg/dL · HbA1c 8.4% resulted today High
I10
Essential hypertension Problem list since 2023 · Telmisartan 40mg on the current list High
Z79.84
Long-term use of oral hypoglycaemic drugs Metformin dispensed continuously since Mar 2024 High
N18.30
Chronic kidney disease, stage 3 unspecified From an outside KFT — eGFR 44, Yashoda Hospital, 14 Mar 2026, via ABHA. This code would be new on her problem list, entered here on the strength of another hospital's single result. Needs you
D64.9
Anaemia, unspecified O: Hb 11.2 g/dL, falling from 12.8. No cause is documented anywhere in her record, and HOS will not pick one — a coded cause would be a diagnosis. Low
Not coded: her 12-day Telmisartan gap and the unresolved lipid order. Neither has an ICD representation — they affect her care, not her claim.
✦Why bother, on a self-pay OPD visit that bills nothing today? Three reasons, and none of them is money this morning. Her problem list becomes correct. Her next admission's package selection becomes defensible against a payer — her Aug 2023 admission was coded E11.65 + E86.0 and cleared first time. And the hospital can answer "how many CKD patients do you have" without a chart review. Codes are stored beside your own local text from the first visit because retro-coding a year of free text is not feasible, and India is moving toward package-based schemes where coding accuracy is existential.

How the coding is holding up

Measured against the codes you actually signed, not against itself
3 min 20 s
You used to spend this per note looking codes up
25 s
You now spend this checking a proposal — your last 40 notes
6 of 240
Proposed codes that were wrong across those 40 notes. You caught all six. This is the number that matters more than the time
1
Systematic error found: it over-proposed E11.9 where E11.65 was right. Corrected 09 Jul; not seen since

Refer Lakshmi Devi — who is this going to?

Pick the recipient first. The letter and its attachments are assembled for that specialty, not from a generic template.
GREEN · drafting
Specialty
No recipient chosen yet — nothing has been drafted.

Referrals out — last 30 days

Who you sent, and whether anything came back
2 with no reply
ReplyPatientToSentReason
None · 21 daysG. Ramulu T-21 todayOphthalmology · Sankara Eye27 JunRetinal screening, diabetic, never done
None · 9 daysP. SaralaCardiology · Care Hospital09 JulExertional chest pain on a treadmill test done here
ReceivedM. Yesu RatnamNephrology · Yashoda02 JuleGFR 38 on a KFT done here
ReceivedSk. NazeerEndocrinology · Apollo24 JunPoorly controlled T2DM on three agents
The two open rows are the reason this list exists. HOS knows a letter left and no reply arrived; it will not chase a consultant on your behalf, but it will keep saying so. G. Ramulu is in your queue today at T-21 — the referral he never got an answer to is for the same screening Lakshmi Devi has also never had.

Lakshmi Devi, 58F — what her record actually contains

T2DM since Mar 2022 · HTN since Jan 2023 · diabetic neuropathy since Nov 2025 · in Room 3 since 10:41 AM
9 facts · 4 sources · 2 hospitals
Value nowWhat it was beforeWho produced itWhen
eGFR 44 mL/min/1.73m²
problem list here reads CKD stage 3a, entered from this result
62 · Jan 2025Yashoda Hospital · via ABHA14 Mar 2026
Metformin 1000mg BD500mg BD from Mar 2022this hospital · Rxsince Mar 2024
HbA1c 8.4%8.1% · 21 Jun  ·  7.1% · 2022this hospital · labtoday
Hb 11.2 g/dL, falling12.8Yashoda CBC · via ABHA14 Mar 2026
Lipid profile — no resultno sample ever loggedthis hospital · order openordered 21 Jun · 27 days
Telmisartan 40mg OD — 12-day gapcollected on time in every other monththis hospital · pharmacy07–19 May 2026
Pregabalin 75mg HS—this hospital · Rxsince Nov 2025
Allergies: Penicillin, Sulfacontradicted by an outside Rx — see belowthis hospital · allergy flag2022
Weight 68 kg68 kgthis hospital · every visit4 years, unchanged

✦ Convene the panel

Three lenses read her record separately and do not see each other's answers. Expect them to disagree — that is what you are here for.
AMBER · reference & consistency
✦Each lens may do exactly three things: state a value that is in her record, quote a constraint published somewhere citable, or ask you a question. None of them may tell you what to do — and the panel has no chairman who resolves them.

✦ Ask the panel

Follow-up questions get evidence, with the record it came from named. They do not get advice — try asking for some.
AMBER · evidence only
✦ The three lenses stay available after they report. Ask any of them for the evidence behind a line — or ask the panel what to do, and watch what happens.
SA
Sahayaka
Your personal assistant · trained on your 412 approved notes and nothing else · never leaves this hospital's database
GREEN · drafts & formats
9
Preferences learned in 30 days
7
Currently running
2
You switched off — and it stayed off
6
Corrections it stopped needing you to make

What it has learned about you — showing the 6 with the biggest effect

You always add a renal-function note for older patients on Metformin ON
Learned from 6 corrections you made in 30 days. The same six appear in Administration → AI Quality as a fix in the owner's queue — this is the doctor's side of it.
Now does automatically: every draft for a patient over 55 on Metformin carries your renal line and cites the eGFR it read, with its date and which hospital produced it. It still does not pick a dose, and it never will.
Your Telugu counselling register is spoken, not formal ON
Learned from 14 of your own Telugu lines rewritten in 30 days. You say షుగర్, never మధుమేహం; you use the respectful plural throughout; you never write a sentence a patient would have to read twice.
Now does automatically: drafts every patient-facing Telugu line in your spoken register, and flags any sentence longer than 14 words for you to shorten.
Your T2DM order set, and where you draw the line ON
Learned from 142 uses of your T2DM follow-up template and 744 T2DM visits recorded here in 12 months.
Now does automatically: pre-selects only the labs you order in more than half of these visits, and shows you the ones it left out rather than hiding them. It will not tick a box on your behalf that you tick less than half the time.
You dictate vitals before symptoms ON
Learned from the order of speech in 412 transcripts. Most doctors here do the reverse.
Now does automatically: lays the draft out in the order you spoke it, so reading it back feels like re-reading your own sentence rather than proof-reading someone else's. This is why your median approval time is 38 seconds.
Shortening your Subjective section to two lines OFF
You switched this off on 04 Jul, after a shortened S dropped a patient's mention of night sweats. It has not been attempted since, and it will not be re-learned unless you turn it back on.
What it would do: compress S to the two most clinically loaded sentences. Currently doing nothing.
Adding "no red flags" to fever notes OFF
You switched this off on 11 Jul. It is not a phrase you use, and a phrase you did not write should not appear over your signature.
What it would do: append a negative-findings line to febrile-illness notes. Currently doing nothing.
Resets all 9 to off. Your notes stay; only the personalisation goes.

What it will never learn to do

It will not choose a drug. It will not choose a dose. It will not write an assessment you did not say, and it will not write a plan at all. Those are RED-tier actions under the CDSCO Software-as-a-Medical-Device framework — a licence we do not hold — so they are not features that are switched off. They are not built. If it ever appears to do one of them, that is a bug, and the button below reports it as one.
✦Where it learns from, precisely: your 412 approved transcripts, the edits you made before approving them, your template usage, and your dictation order. Not other doctors here. Not other hospitals. Not anything you rejected without correcting. Nothing it learns is visible to management — the owner sees counts of corrections, never your notes.
SN
Sandarbha
General Medicine reference · depth in the specialty, not opinions about your patient · every line traces to a citation
AMBER · reference

What this specialty actually orders here

Your rateItemThis hospitalT2DM visits
97%HbA1c94%744
76%Lipid profile71%744
61%Kidney function test58%744
34%Urine albumin (ACR) least ordered31%744
12%Retinal screening referral14%744
744 T2DM follow-up visits recorded in this hospital in the last 12 months. This is a count of what was ordered, not a statement about what should be. The urine ACR is the item this specialty leaves out most often, here and in the published Indian audit literature alike.

What the season is doing in Kukatpally

Monsoon window. Ward-level dengue notifications for Kukatpally: 42 this fortnight against 11 in the same fortnight of 2025 — Telangana IDSP weekly bulletin, week 28 of 2026. In this hospital's own OPD, presentations recorded as "fever, 3 days or more" are running at 3.1× the May baseline: 71 in the last 14 days against 23 in a comparable May fortnight.

Mohd. Irfan (T-14) is in your queue now, recorded as fever, day 3. That is a record fact, and the bulletin count above is a published number. They are on the same screen because you asked for the specialty picture. Sandarbha is not putting them together for you — connecting a surveillance count to a named patient is a screening judgement, which is not something it is licensed to make.

TB continuation phase. Six patients under your care are in the continuation phase of a six-month regimen. The NTEP treatment card schedules a follow-up sputum smear at the end of month five. Four of the six reach the end of month five this week, and none of the four has a smear recorded. That is a schedule published by the Central TB Division held against your own records.

How to say the difficult things in Telugu

A three-month average, not today's sugar
మీ మూడు నెలల షుగర్ సగటు 8.4 ఉంది. లక్ష్యం 7 కంటే తక్కువ. ఇది ఒక్క రోజు షుగర్ కాదు — మూడు నెలల సగటు.
Your three-month sugar average is 8.4. The target is under 7. This is not one day's sugar — it is a three-month average.
A kidney number from another hospital
వేరే ఆసుపత్రిలో చేసిన కిడ్నీ పరీక్షలో ఒక సంఖ్య తక్కువగా వచ్చింది. దాన్ని ఇక్కడ మళ్లీ చూడాలి. భయపడాల్సిన అవసరం లేదు, కానీ వదిలేయకూడదు.
A kidney test done at another hospital came back with one low number. We should look at it again here. There is no need to be frightened, but it should not be left alone.
Asking about a refill gap without accusing
మే నెలలో పన్నెండు రోజులు మాత్రలు తీసుకోలేదని రికార్డులో ఉంది. మర్చిపోయారా, లేక డబ్బు ఇబ్బందా? నిజం చెప్పండి — తప్పు లేదు.
The record shows twelve days in May without tablets. Did you forget, or was money difficult? Tell me the truth — there is no blame in it.
Why a BP tablet is not a painkiller
బీపీ మాత్ర రోజూ వేసుకోవాలి. నొప్పి లేకపోయినా వేసుకోవాలి. అది నొప్పి మాత్ర కాదు.
The BP tablet has to be taken every day, even when there is no pain. It is not a painkiller.
Phrasing only. Sandarbha writes the words; you choose the message. The third line is the question the cost lens raised in Case Discussion, in the language she would answer it in.

Documentation templates for this specialty

T2DM with reduced eGFR — review note

Structure only: S, O, A, and an empty P. Prompts for the eGFR's date and source.
Used 31 times
GREEN

Nephrology referral — single unrepeated eGFR

Assembles both eGFR values with dates and sources, the ACR gap, and the current drug list.
Used 9 times
GREEN

Monsoon febrile illness — day-of-illness note

Forces an explicit day of illness, which your notes already do 11 times out of 11.
Used 88 times
GREEN

TB continuation-phase visit — NTEP card fields

Mirrors the NTEP treatment card: month, weight band, smear schedule, adherence entry.
Used 46 times
GREEN

Ask Sandarbha a reference question

✦ It answers "what does the reference say" and "what does this hospital's data show". It will not answer "what is wrong with this patient".
✦Citations, because a reference agent without them is just a confident stranger. Sources on this card: Telangana IDSP weekly bulletin, week 28/2026 · Central TB Division NTEP technical and operational guidelines · licensed prescribing information held in this hospital's formulary · this hospital's own 5,880 General Medicine OPD visits and 744 T2DM follow-ups over 12 months · published Indian diabetes-care audit literature. Nothing on this card is recalled from a model's memory, and every line above names which of those it came from.