HowdyHire

Inspiration

As college students (especially freshmen) we know how intimidating it can be to apply for on or off campus jobs without any prior work experience or interview practice. Many students have never had a formal job interview before, which can make the process feel unfamiliar, stressful, and overwhelming.

Although our campus career center offers valuable resources, accessing those services can be inconvenient for students who live off campus, have packed schedules, or need practice outside of regular office hours.
Many existing online mock interview tools rely on simply recording answers to preset questions. While helpful, these tools often fail to recreate the pressure, back-and-forth conversation, and emotional intensity of a real interview.

We wanted to build a platform that makes interview practice more accessible, realistic, and personalized for students preparing to enter the workforce.

What it does

Our website simulates job interviews based on the user’s desired role. Users can input their resume and key background details, such as past jobs, projects, leadership experience, skills, and career interests, allowing the system to generate more personalized and relevant interview questions.
The platform creates questions connected to the user’s actual experiences and the position they are preparing for. This makes the practice more realistic than generic mock interview questions and helps students prepare for the types of questions they may actually face.

How we built it

HowdyHire pairs real-time voice AI with custom browser tooling for authentic interview practice:

Chrome Extension: Manifest V3 extension that scrapes job listings from HireAggies or Indeed in one click to populate role context.
ElevenLabs Voice Studio: Powers hyper-realistic, low-latency voice personas (Easy Aggie Alum, Moderate Morgan, and Tough Alex) for immersive spoken reps.
Serverless Backend & API Routes: Built on Next.js API Routes and Node.js to handle asynchronous AI prompt orchestration, secret management, and payload passing securely.
Real-Time STAR Analytics Engine: Server-side and client-side processing that parses spoken audio live to track pacing (WPM), filler words, and STAR framework coverage.
Next.js & Aggie Design: Built with Next.js 14, TypeScript, and Tailwind CSS using a Maroon & Cream (#500000 / #FDFBF7) palette and circular Reveille mascot UI.
Session Persistence: StudioStepper state architecture lets users instantly "Run It Back" on any question with full context saved to the Aggie Dashboard.

Challenges we ran into

Designing a user interface that felt professional and easy to navigate while still being visually engaging. Since our target audience is students preparing for real job opportunities, we wanted the website to feel polished and trustworthy without making it intimidating.
Determining the most effective format for feedback was also tough. We wanted users to receive detailed insights without overwhelming them with too much information at once. Finding the right balance between clear scoring, constructive criticism, and practical next steps was an important part of our development process.
We also explored how to connect the system to real job listings in the Aggies for Hire database. Integrating actual job data would make the platform even more relevant by allowing students to practice for specific roles they are interested in applying to.

Accomplishments that we're proud of

Building a Chrome Extension that allows users to pull job listings directly from platforms like HireAggies or Indeed, making interview practice more relevant to real opportunities.
Implementation of asynchronous AI prompt orchestration through our serverless backend and Next.js API routes, allowing the platform to securely manage user data, role context, and AI-generated interview questions in real time.
Developing our real-time STAR analytics engine, which evaluates spoken responses as the interview happens by tracking pacing, filler words, and coverage of the Situation, Task, Action, and Result framework.

What we learned

The way students practice for interviews can significantly impact their confidence and performance. Simply recording answers to interview questions does not fully prepare students for the conversational nature and stress of a real interview. A simulated interview environment can better help users practice thinking on their feet, responding naturally, and managing pressure.
Structured feedback is important. Job seekers need more than a general sense of whether an answer was “good” or “bad.” By using a clear rubric like the STAR method, users can understand exactly what their responses are missing and how to improve. This type of targeted feedback is especially valuable for students who may not have prior interview experience.
We each learned to build on our foundational knowledge of Github; Rohanna gained familiarity with Github's layout and the necessity of API keys, Sahasra learned how to use Figma to create a demo for the website, and Sragvi learned how to manage complex fullstack state across APIs and browser extensions.

What's next for HowdyHire

Expanding the platform by incorporating camera capture and visual analysis. This would allow the system to provide feedback on nonverbal communication, including eye contact, posture, body language, and overall presentation.
Developing the platform’s connection to real job postings, such as those in the Aggies for Hire database, optimizing how tailored the questions are to specific opportunities.

Ultimately, our goal is to make interview preparation more accessible, realistic, and useful for students as they begin applying for jobs and internships.