Inspiration
Searching for jobs can feel exhausting and random: endlessly filling out applications, cold-messaging recruiters, and still feeling like you're shouting into the void. We thought: what if getting a job felt as easy as saying “Open Sesame”? We wanted to build a platform that automates job searching and networking, so talented people spend more time getting interviews, not stuck doing busy work.
What it does
Open Sesame lets users:
- Select their experience level, industry preferences, and job preferences.
- Upload a resume.
- Automatically apply to job postings online that match their criteria.
- Send personalized LinkedIn connection requests to recruiters, using AI to craft unique messages based on the user’s profile and goals.
All users have to do is say "Open Sesame" — and we do the rest.
How we built it
We built the program from scratch using:
- A Python (Flask) backend that serves the frontend and handles form submissions.
- A React + HTML + JavaScript frontend for a smooth and simple user experience.
- Gemini-powered Browser-Use to automate application submissions and sending LinkedIn requests.
Key technical highlights:
- Built the interface between user data and browser automation.
- Managed browser workflows to run in parallel efficiently, maximizing speed and minimizing crashes.
- Integrated resume parsing and keyword extraction to better match jobs and generate personalized recruiter messages.
- Created fallback and retry logic in case a connection or application attempt fails.
Challenges we ran into
- Speed and concurrency: Managing multiple simultaneous job applications without being flagged by job boards or causing rate-limiting errors.
- Browser automation reliability: Handling errors, login issues, and multi-tab browsing without breaking the flow.
- Personalization quality: Making sure AI-generated LinkedIn messages felt human, not robotic.
Accomplishments that we're proud of
- Fully functional end-to-end demo: from user inputs ➔ to real job applications ➔ to real LinkedIn outreach.
- Making the system robust enough to handle retries and parallel processing.
- Designing a UI that feels simple and magical — hiding complex backend automation under the hood.
What's next for OpenSesame.Work
- Learning from outreach: Analyze recruiter response rates to continuously improve and personalize future connection messages using reinforcement learning.
- Enhancing matchmaking: Smarter AI that better understands user career goals and job descriptions.
- Building a dashboard: Let users track applied jobs, recruiter responses, and manage their networking pipeline.
Our dream is simple: Job hunting should feel magical — not miserable.
Log in or sign up for Devpost to join the conversation.