CareerForge-AI

The Ultimate AI Career Copilot powered by Google Gemini 3 Pro.

Comment

Inspiration

The seed for CareerForge-AI was planted during a period of intense frustration for our team leader, Hanson He. As he navigated the grueling process of job applications and technical interviews, Hanson realized how inefficient and manual the process remained, even in the age of AI. He found himself spending hours tailoring resumes and lacked a high-quality, real-time mock interview partner that could truly simulate the "deep thinking" required for top-tier roles.

Driven by this personal struggle, Hanson He led the development of a solution that could transform the job hunt from a source of stress into a streamlined strategy. He envisioned a system where AI doesn't just "fix" a resume, but acts as a strategic career partner that understands the deep semantic alignment between a candidate's experience and a company's vision.

What it does

CareerForge-AI is an advanced career co-pilot powered by the Gemini 3 API. It acts as an end-to-end agent for job seekers:

  • Reasoning-Based Resume Optimization: Instead of simple keyword stuffing, it uses Gemini's deep reasoning to restructure experiences to match the specific logic of a Job Description (JD).
  • Agentic Mock Interviews: A multimodal interview simulator that provides real-time, critical feedback on both technical accuracy and communication style.
  • Strategic Match Scoring: It calculates the "Match Probability" for any given role, helping users prioritize their efforts.

How we built it

The project was architected and driven almost entirely by Hanson He. He spearheaded the technical roadmap, from initial concept to the final integration of the Gemini 3 API.

  • Core Engine: We utilized Gemini 3 Pro for its superior reasoning and agentic workflows, and Gemini 3 Flash for low-latency interview interactions.
  • Frontend/Backend: Built with a modern stack including Streamlit and Python, ensuring a fast and secure user experience.
  • Algorithmic Logic: Under Hanson's guidance, we implemented a custom scoring system to evaluate candidate-job fit.

Challenges we ran into

The biggest technical hurdle was ensuring the AI maintained high "grounding" accuracy—preventing it from hallucinating experiences while trying to be persuasive. Hanson He solved this by designing a "Master Profile" verification layer. The AI is restricted to a strict set of user-provided facts, ensuring that every resume generated is 100% authentic while remaining highly optimized.

Accomplishments that we're proud of

We are incredibly proud of the Agentic Interviewer. Under the leadership of Hanson, we built a system that doesn't just ask static questions but "listens" and asks complex follow-up questions based on the user's previous answers, mimicking the pressure of a real-world interview.

What we learned

We learned that the true power of Gemini 3 lies in its agentic behavior. It is no longer just about generating text; it’s about the model’s ability to plan and reason through a user's career path. Hanson’s leadership taught the team how to move beyond simple prompts to building complex, self-correcting AI workflows.

What's next for CareerForge-AI

We plan to expand the platform to include multimodal "Social Proof" generation, helping users create AI-driven portfolios and personalized networking messages. Our goal is to make Hanson He's vision of a frictionless job market a reality for every professional.

Built With

Share this project:

Updates