Tell me a story you've never told publicly.
About the time a doctor asked you for a feature and you didn't build it.
That sounds like an accusation.
It's a bit of one.
Hello, I'm Tomasz Szuster.
I built the mojApteczka app.
And this story is about how, sometimes,
the most honest answer to a specialist's request
is to build something other than what they asked for.
And then explain why.
Start at the beginning, then.
What did they say?
The doctor suggested something simple,
that the owner of the medicine cabinet
should be able to flag a medicine they're allergic to.
One field, you tick it, and that's that.
And that sounds sensible.
It sounds sensible,
and that's exactly how I'd have built it if I hadn't stopped for a day.
What happened in that day?
I asked myself what happens when a different box turns up in the cabinet.
Meaning?
You flag the allergy on one particular pack.
A month later, you buy a medicine from a different company.
Different name, different box.
Inside, the same substance.
And the app says nothing.
Says nothing.
Because the flag was sitting on that box,
not on the substance.
And that's precisely the case where the warning was needed most.
So the feature would exist.
It'd look good on a list of features.
And it'd stay quiet at exactly the moment it ought to speak up.
That's worse than not having it at all,
because it gives you a false sense that somebody's keeping an eye out.
So what did you build instead?
I moved the allergy off the pack and onto the active substance.
And I added a second dimension that wasn't in the original request at all.
The person.
The person?
Because a medicine cabinet at home doesn't belong to one person.
An allergy belongs to a particular human being.
You, me, a child, grandma.
The warning has to know who it's about.
Otherwise it's just noise.
That seems obvious when you put it like that.
It seems obvious afterwards.
It wasn't in the request,
because the doctor was thinking about the patient sitting across the desk from them.
I'm dealing with a drawer that five people use.
By the way, everything we're talking about here really is a feature in the mojApteczka app.
This isn't an advert, just context to be clear.
Right.
But now I want to hear the other side,
because you sound as though you've solved it.
I haven't solved the whole problem.
And that's actually the more important half of this episode.
Go on.
There are three things this feature doesn't do,
and all three are written down in the design documentation
as deliberate limitations of the current version.
As boundaries, you have to say out loud.
The first?
There's no cross-reactivity.
In plain terms,
if you enter an allergy to one substance,
the app will not automatically warn you about another related one,
unless you enter that one separately as well.
Hang on.
That's fairly serious.
It is serious.
Which is why I'm saying it on a podcast,
and not only in the documentation.
Drug classes are clinical knowledge,
and I'm not going to build that on my own out of a database of names.
This is exactly the point where a conversation with the pharmacist
has no substitute.
The second thing?
The substance matching is
heuristic,
which means that,
with an unusual spelling,
the warning might not appear at all,
and sometimes it might appear when it needn't.
So it can both miss things
and give you a fright.
Both.
And again,
I'd rather say so
than pretend to a precision I haven't got.
And the third?
The third is the most contentious.
The app warns,
but it never blocks.
It doesn't block.
So I can see the warning
and still hand that medicine out?
You can.
It's deliberately built that way.
Explain that,
because at first glance it sounds irresponsible.
Because I don't know what situation you're in.
Maybe the doctor decided that deliberately.
Maybe the entry is out of date.
Maybe it was a slip when it was typed in.
An app that blocks
is making a medical decision on a person's behalf,
and that's not something it's allowed to do.
So blocking would look safer,
but it'd be crossing a line.
Exactly.
A warning says,
check this.
A block says,
you mustn't.
Only one of those is mine to say.
You said those are documented limitations,
but was there a bug that actually happened?
There was,
and it's an uncomfortable one.
With a feature that works per person,
the most dangerous class of bug
is one person's data leaking onto another.
Meaning,
my allergy shows up against Leon's medicine.
Exactly that.
And it isn't theoretical.
It's the class of mistake
that had to be closed off with tests of its own.
Because that kind of bug is sly.
The app still works.
It still shows you something.
It just attaches it to the wrong person.
And nobody notices,
because the warning looks perfectly normal.
Completely normal.
Which is why,
with family features,
what interests me least
is whether something displays.
What interests me
is whether it displays
to the right person.
Now the question
we probably should have started with.
How often do medicine mix-ups
happen in Polish homes?
I don't know.
What do you mean,
you don't know?
I don't know of a reliable Polish figure.
And I want to be precise here
because these are two different things.
Which two?
The first is an adverse reaction to a medicine.
The medicine did somebody harm.
There's an official route for that.
A patient,
a carer,
or a pharmacist can report it.
That works.
And it's good that it works.
And the second?
The second is a plain mix-up.
Taking the wrong medicine.
A double dose,
because two people both gave it.
Giving a child
something meant for adults.
Here the medicine did exactly
what it was supposed to.
It just went to the wrong place.
And you couldn't find
that collected anywhere?
I haven't found a public,
systematic picture
of what goes on
at the drawer at home.
I could be wrong,
and I'd be glad
if somebody showed me
that such a picture exists.
So you haven't got a number?
I haven't.
And I could do the thing
half this industry does,
take a statistic
from another country,
multiply it by the population,
and present it as a Polish one.
It'd look very convincing
on a slide.
But?
But it'd be made up.
A different health system,
a different way of counting,
different habits.
It doesn't scale by head count.
So you're building
a safety feature,
and you don't know
how much of the problem
you're solving.
I don't.
And honestly,
that's the most frustrating thing
about this job.
I don't know how many times
a warning like that
has genuinely helped somebody,
because the help
consists of something
not happening.
And I can't see that
measured anywhere.
So,
what do you do about it?
You can't set up
a register on your own.
I can't,
and I shouldn't.
But I can say out loud
that the number
is missing.
Because as long as there's
no public,
systematic measurement,
it's easy to treat this
as a problem nobody can see.
And things nobody can see
rarely get programs
and money.
And that's the ask.
That's the whole ask.
Not,
buy my app.
Just,
let's count it.
Anyone.
If only so it turns out
the problem is smaller
than I think.
Would you be alright
with that result?
I'd be alright
with any result
that's true.
I'm not attached
to the problem being big.
I'm attached
to it being known.
I'll add mine,
in ordinary human terms.
In our house,
the medicine cabinet
isn't one shelf
belonging to one person.
Leon's nine,
Basha's five,
and those are
completely different medicines.
And then there's my mum,
who sometimes stays over.
And that's the drawer
I was talking about.
Not the patient,
but four people
and one box
for the lot.
Which is why,
if I've got any doubt at all,
I take the pack
and go to the chemists.
And that's the right answer.
The app's job
is to remind you
there's something
worth asking about.
The answer
belongs to the pharmacist.
If you'd like to give it a go,
you'll find the link
to the mojApteczka app
in the episode description.
The app is free
and your medicine cabinet
is there offline too.
And if you enjoyed this episode,
share it with someone
who sorts out the medicines
for the whole family.
One more thing
before we go.
This podcast
is made with the help
of artificial intelligence.
I wrote the dialogue myself,
but the voices you're hearing
aren't ours.
They're AI-generated,
synthetic voices.
We're telling you this
straight out
because we think
it should be the standard
for any podcast
made this way.
Full details
are in the show notes
and in every episode description.
Until next time.
Until next time.