Inspiration

Every student I know has the same story. They apply to an internship, get excited and then find out three weeks in that the stipend from the post never shows up. Or worse, the internship was never real to begin with. Research backs this up too. Over 60% of internships in fields like IT, media and finance in Pakistan are unpaid or pay next to nothing, and the government has had to publicly warn students about fake internship postings multiple times this year. What struck me is that every student learns this the hard way on their own and none of that experience ever reaches the next person walking into the same trap. InternIQ came from wanting to turn everyone's individual bad experience into something the next student can actually use.

What it does

InternIQ helps students avoid two problems before they commit their time. It scans an internship or job posting's text and flags common scam patterns like upfront fees, vague company details, personal email addresses instead of company ones and unrealistic stipend promises. It also gives students access to anonymous crowdsourced reviews from students who already interned at that company, showing whether the pay was honored, whether the work was real and whether past interns would recommend it.

How we built it

I started from the two failure points students actually described. Getting scammed outright and getting exploited by real but dishonest companies. For the scam detection side, I looked into common red flag patterns in fraudulent job and internship postings such as fee requests, generic email domains and vague company identity to figure out what the classifier needs to catch. For the transparency side, I built the review system around the specific questions that actually matter to a student deciding whether to apply. Was the pay honored? Was the work real? Would they go back? these felt far more useful than a generic star rating that does not actually tell you anything.

Challenges we ran into

The hardest part was keeping the review system genuinely useful without turning it into a place for unfair pile ones. A bad review needs to be specific and honest instead of just venting. Otherwise companies and future students relying on it won't trust it. I also had to think carefully about the scam classifier's mistakes. A false positive flags a real legitimate posting as a scam, and if this happens too often students stop trusting the tool completely and start ignoring even its correct warnings. That's why the detection needs to stay conservative and explain its reasoning instead of just showing a red flag with no context. A false negative is far more dangerous. It means the tool misses an actual scam and tells the student the posting looks fine, so they apply with their guard down and get scammed anyway. This isn't just the tool failing to help. It's the tool actively giving false confidence right before someone gets hurt, which is a much worse outcome than being overly cautious.

Accomplishments that we're proud of

Turning something every student silently experiences and never talks about (being scammed or exploited by an internship) into a structured shared resource that actually changes what the next applicant decides to do. It's more than just another place to complain.

What we learned

I went in thinking the main problem was outright scams. But the research showed the bigger, more common harm is legal but dishonest companies exploiting the fact that nobody shares information. The unpaid internship problem affects far more students than outright fraud does. This changed how much weight the review and transparency side of the idea deserved compared to the scam detection side.

What's next for InternIQ

Partner with university career offices to seed the review database so it has real coverage before students need to rely on it
Add a simple report a scam posting flow that feeds back into improving the classifier over time Explore verified reviews tied to proof of internship (like an offer letter) to keep the database trustworthy as it grows