Taking on new projects
Huzaifa Tahir
After-hours and missed-call coverage for US local service businesses. HIPAA-ready for medical practices.
Most health tech is sold to investors and built by engineers. The people who actually use it all day are doctors, nurses, front desk staff and billers. And almost nobody builds for them.
So that is who I build for. Tools that write up notes, chase claims, fill empty appointment slots and answer the phone when nobody can. Then I write about how they work, in words you do not need a tech background to follow.
If something in your clinic eats hours every week, I can probably help. And if software is the wrong answer, I will tell you that too. It saves us both time.
AI that helps your team decide faster. It suggests, flags and drafts. A person always makes the call.
Patient records, booking, telehealth and billing tools. Built to fit the systems you already run.
Reminders, follow-ups, refills and claim chasing. The jobs nobody has time to do by hand.
Articles for nurses, doctors and clinic managers. No code, no jargon, nothing to look up.
The tools behind the work. Here is what each one actually does for a clinic.
The AI I use most. Good at reading messy notes and writing clear, careful drafts.
The other big AI provider. Useful for voice, images and scanned paperwork.
Lets AI answer from your policies and records instead of guessing. Called RAG.
Finds the right document by meaning, not exact words. Staff stop hunting through folders.
Not just chat. It books, sends, files and follows up, then hands you anything unclear.
Answers WhatsApp, SMS and calls after hours. Urgent cases go straight to a person.
See it working
A patient messages the clinic at 9pm. The agent answers, books and confirms on its own. Then it hits something urgent, and hands straight to a human.
Chest tightness is caught by a fixed rule written into the code, not by the AI deciding it sounds serious. The model can add to that list. It can never take anything off it. The patient is told to call 911 and a human is paged. Both, every time.
Scripted replay of a real workflow. No live model is called from this page, and no real patient information appears in it.
Find AI projects built for your industry. Filter by niche to see exactly how AI solves problems in your sector.
Showing 11 AI projects
Listens to the visit and writes the note, the codes and the follow-up instructions. Gives a doctor about two hours back a day.
Drafts the doctor notes from the visit and suggests the billing codes. A doctor reads it, fixes anything wrong, and signs.
Suggests billing codes from the notes, points out what the note is missing, and warns you which claims are likely to bounce.
Reminders patients can reply to, and a waitlist that fills the empty slot within minutes. Took one practice from 18% no-shows to 7%.
Builds the insurance approval request from the notes, checks it against the rules before sending, and tracks it. Nine days down to two.
Answers the calls and messages your front desk cannot, books appointments, and gets anything urgent to a person within a minute.
Handles the routine insurance claims end to end and sends the tricky ones to a person, cutting a two-week wait down to two days.
Sorts incoming insurance claims, checks them against the policy, and drafts the reply letter for an adjuster to check and send.
Regular check-ins with staff that spot who is struggling early, and point them to real help instead of a leaflet.
Work out what missed calls, no-shows and slow follow-up are really costing you. One calculator per kind of business. No sign-up, no email needed, and nothing you type leaves your browser.
How AI and modern software are transforming medicine. For doctors, patients, and caregivers
A missed appointment is a slot you already paid for. Here is the reminder cadence, the waitlist and the risk flagging that reduce no-shows, and what to automate first.
Sepsis kills more Americans than any other in-hospital condition, and many of the AI tools built to catch it early are backfiring by burying clinicians in false alarms. Here is why sepsis prediction is one of the hardest problems in clinical AI, what went wrong with widely deployed models, and the engineering and workflow approach that is proven to fix it.
A deep look at the state of US healthcare software in 2026, where Epic, Oracle Health, athenahealth, and Veradigm fall short, what new CMS interoperability rules demand, and how an AI-native, API-first approach can out-compete the incumbents.
Tell me what is eating your team's time. I will tell you straight whether software can fix it.
You do not need a spec or a budget worked out. A few lines about the problem is enough. If I am not the right fit, I will say so and point you somewhere better.
Book a free 30-minute callPick a time that suits you. No pitch, no obligation. Or email me below if that is easier.
Before any patient data is involved. If a vendor in the chain will not sign one, I will tell you and we design around it.
Nothing I build writes to a record or reaches a patient without a person reviewing it. Red flags escalate by fixed rules, never model judgement.
Usually the same day. I work US hours for calls, and you will deal with me directly, not an account manager.
Response Time
Usually within 24 hours
Book a Meeting
Free 30-minute call, no pitch
