yes, it should be "3indaki"
to a male, it would be "3indaka".
So, the basic word is (3inda) - meaning at - and then the pronoun is attached according to the gender; (-ki) to a female and (-ka) to a male.
In dialects nowadays, they somehow flip the spelling and it becomes (-ik) to a female and (-ak) to a male.
But colloquially they do say "3indak" (masculine) and "3indik" (feminine), right? ((...for standard 3inda-ka and 3inda-ki, respectively...))
So I guess the orthography indicates the latter (ie 3indik) but the Text-to-Speech then reads the kaf character in isolation, which creates this unreal blend *3indi-ka, from what should have been just 3indik. That's my analysis of this error; correct me if I'm wrong.
Yeah, I guess this is what's going on here. In this course it's like more than 50% of the audio are wrong. I literally stopped correcting. It's a clash between dialects and proper standard Arabic and the speech machine is not helping as well as you can see.
Just a note: the last name should be Sámiyah, not Sámiyatan.
Well, the problem with Duolingo (in all languages) is that it translates literally sometimes and by meaning other times.
Anyway, by general meaning, yes your sentence is equivalent to the Arabic one. Literally though it is different because it contains the word (any). That would be: هل عندك أي مشروب يا سامية؟
and since the audio and the typing is wrong up there I should probably put it in a proper way: هَلْ عِندَكِ مَشْروبٌ يا سامية؟ (hal 3indaki mašrúbun yá sámiyah).
Any = أي (which is also used as the interrogative particle "which?").