Inspiration
What inspired me is that when we prepare for placements we have to keep switching between many platforms. One website for the resume, another for aptitude, another for coding questions, and for the HR interview we practice with a friend or nobody at all. And there is no proper place where we can showcase our skills, test them and see how ready we are. So I thought of building one platform which is a single stop for everything a student needs for placement preparation.
What it does
PlacementAI is an AI Resume Builder with Placement Preparation. A student chooses a job role and a template and the AI writes the resume summary and career objective from their details. Then they can check the resume in the ATS checker, see the score and the missing keywords and also what to improve.
They can practice aptitude tests (quantitative, logical reasoning and verbal) with a timer and automatic scoring, practice technical questions in Java, Python, C++, SQL and React, and give an AI HR mock interview where the AI gives feedback on their answer.
All of these are connected on one dashboard with a placement readiness score, progress and recent activity, so the student knows which part needs more work. The history of every round is kept as context for that student, so the questions and feedback are based on their own work and not the same for everyone. It is also a Progressive Web Application so it can be installed on a phone.
How we built it
The frontend is made with React js and Vite. The backend is Node js and Express. For login and the database I used Firebase (Authentication and Firestore), and the AI features like resume writing, ATS analysis and HR interview feedback use the Grok API. The website is deployed on Vercel.
I created each module separately (resume builder with many templates, ATS checker, aptitude, technical, HR interview, dashboard) and then connected all of them to one placement readiness score, so the dashboard shows the real results of the student and not fixed numbers.
Challenges we ran into
The first challenge was making the insights real. In the beginning the dashboard and results looked good but many things were static, so every student saw the same type of data. I had to change it so the scores, feedback and questions come from each student's own work, and that was harder than just building the screens.
The second challenge was personalized data for different users. Two students with different branch, skills and target role should get a different experience, so I had to store each student's history and use it as context instead of treating every test as a new start.
The third challenge is the risk that AI is developing so fast and can do almost everything now. A student can ask any chatbot for a resume or a mock question, so just using AI is not a differentiator. What we built is different things combined in one place. Aptitude, technical and HR rounds, the resume and a road map are connected, and the history of all of them is kept as context. A general AI chat does not have this full picture of one student at the same time.
There were also normal bugs on the way. In the aptitude test a failed test was showing passed, wrong answers were not shown with reasons and the same questions came again. Storage was a problem too, because Firebase Storage was not available on the free plan.
Accomplishments that we're proud of
We are proud that PlacementAI is a complete working platform and it is live, not just a design. Resume building, ATS checking, aptitude, technical practice and HR interview are all in one website, and the dashboard connects them into one readiness score. Also that a student can open it on their phone like an app. And the biggest thing is that it solves a problem I faced myself as a student.
What we learned
I learned that this is not an easy task. Building a feature is one thing, but making it useful is another. I had to look at features from the end user's point of view and not only like a developer, because a student may use the platform in a different way than I think. Sometimes something I felt was a good feature was not what a student needed.
I learned and developed a lot while building it. The insights, resume building with many templates, the different test modules and then integrating all of them into one platform. Getting all of them to work together was the hardest part.
I also learned that AI output cannot be trusted blindly. Scores and feedback have to be calculated and checked properly, and the API key and student data have to be kept safe.
And the biggest thing is how genuinely it helps students and users in preparing. A platform with many features is not enough if it does not actually improve their preparation, so every feature should be judged on that.
What's next for PlacementAI
First I want to add a voice based HR interview. Students speak their answers and get feedback on how they said it, like their pace, pauses, filler words and voice confidence, along with the feedback on what they said. I want this feedback to be correct and clear, so students can really act on it and not just get a random score.
Then a company wise road map. A student selects their target company and gets a step by step road map for that company, in multiple languages so students can understand it in the language they are comfortable with.
The most important thing is that I do not want to keep it limited to students only. I want to connect with students and different companies, take feedback from them and improve the platform. If companies can use PlacementAI, then it benefits the users the most, because we can prepare them the way big companies actually need and hire.
After that, personalized onboarding with the student's resume, GitHub and LinkedIn, a real coding compiler with test cases and more resume templates.
Built With
- cloud-firestore
- css
- express.js
- firebase
- firebase-authentication
- grok-api
- javascript
- node.js
- pwa
- react
- tailwindcss
- vite
