CareerBoost AI

Inspiration We were inspired to build an end‑to‑end, AI‑driven experience that covers every step of the job search for students — from discovering roles and submitting resumes, to mock interviews and actionable feedback — so nothing falls through the cracks between preparation and real hiring.

What it does CareerBoost AI is an end‑to‑end, AI‑powered platform for students that helps them find jobs, improve resumes, run mock interviews, and receive structured feedback — all in one place. It’s designed as a multimodal experience with text, voice, and video interview flows to simulate real‑world hiring as closely as possible.

How we built it We built the application with a FastAPI backend and a Tailwind/TypeScript frontend, integrating Tavus for the avatar and AGNTCY for multi‑agent orchestration. We also use RAG and Skills to tailor interview questions and feedback to each role and candidate, and we designed the system to support video, audio, and text streams as part of the interview workflow.

Challenges we ran into Our biggest challenge was synchronizing real‑time data streams to keep the AI avatar’s responses and interview feedback aligned with low latency across multiple modalities.

Accomplishments that we're proud of We’re proud of creating a closed‑loop system that takes a resume and job description, runs a realistic interview, and outputs actionable feedback to help candidates improve quickly. We also built the foundation for a multimodal interview experience that can switch between text, voice, and video without breaking the flow.

What we learned We deepened our understanding of RAG, the impact of Skills‑based prompting, and how to build multimodal pipelines that coordinate audio/video/text in real time. We also strengthened our front‑end architecture skills with Tailwind’s atomic styling and complex state management using React Hooks.

What we created We created a full‑stack, multi‑agent platform that covers job discovery, resume submission, mock interviews, and feedback, with a multimodal interview layer and a scalable backend architecture ready for production‑grade expansion.

What's next for CareerBoost AI Next, we plan to launch “Knowledge Gap Analysis,” automatically generating a personalized 7‑day learning roadmap based on interview performance, and to fully roll out real‑time voice and video interview sessions.