VERIJOIN

main

Inspiration

Students and early-career professionals often receive job or internship offers online and must wait for verification. During this waiting period, they face uncertainty, financial pressure, and a lack of direction. We were inspired by the idea that AI should not only verify documents, but also support users during delays by helping them learn, earn, and prepare instead of staying idle.

What it does

VeriJoin is an AI-powered career trust platform that verifies offer letters and actively guides users while verification is in progress.

Users can:

Upload an offer letter for AI-based verification

Track verification status, confidence, and time metrics

Access learning courses if verification is delayed

Discover trusted part-time job opportunities during the waiting period

Prepare for interviews with targeted guidance

Explore relevant company hiring activity and updates

VeriJoin transforms waiting time into a period of skill growth, income opportunities, and interview readiness.

How we built it

VeriJoin uses an agent-driven architecture powered by Gemini 3:

Gemini 3 API for multimodal offer letter understanding and reasoning

Next.js 15 for the frontend experience

Python (FastAPI) backend for workflow orchestration

Gemini 3 dynamically:

Analyzes offer letters

Detects verification delays

Recommends courses, interview preparation paths, and trusted part-time jobs

Adapts guidance based on role, skills, and waiting time

This makes Gemini central to decision-making, not just content generation

Challenges we ran into

Designing a single flow that combines verification, learning, and job discovery

Keeping the demo public and rule-compliant while still feeling professional

Avoiding feature overload while showcasing real-world usefulness

Ensuring recommendations felt relevant and contextual

We addressed these by prioritizing a smooth end-to-end experience and focusing on user needs during uncertainty.

Accomplishments that we're proud of

Built a public, no-login demo aligned with hackathon rules

Created a system that supports users during verification delays, not just after results

Integrated learning, part-time jobs, and interview prep into one AI-driven flow

Designed VeriJoin to feel empathetic, practical, and career-focused

You said:

Initializing VeriJoin Project Documentation I have completed the

README.md creation with the 'Built With', 'How we built it', and 'Instructions for Judges' sections. I am now updating task.md to reflect this progress.

Files Edited Task

README.md Progress Updates 1 Creating README.md and task.md

2 Writing README.md with detailed tech stack

3 Adding 'Instructions for Judges' to README.md

4 Updating task.md to mark documentation tasks as complete

I have updated the README.md with the "Built With" section featuring the high-end 2026 tech stack, included the "Agentic Vision" pro-tip in the "How we built it" section, and drafted the "Instructions for Judges" based on the features described.

Would you like me to proceed with scaffolding the Next.js 15 application and the Python backend structure as described in the documentation?

README.md Open ok next

Thought for 10s

Scaffolding VeriJoin Application I have completed the scaffolding of the Next.js frontend and Python backend, including applying the initial FinTech styling. I am now updating the task list to reflect this progress.

Files Edited

globals.css Task

.gitignore

main.py

requirements.txt Progress Updates 1 Initializing Next.js 15 frontend

2 Waiting for Next.js initialization to complete

3 Applying FinTech styling to frontend

4 Marking scaffolding tasks as complete in task.md

I have successfully scaffolded the Next.js 15 application in the frontend directory and set up the Python backend in backend. I've also applied the "FinTech" color palette (Deep Navy/Slate) to the frontend.

You can now run the frontend with npm run dev inside the frontend folder.

Shall we proceed to implementing the specific "Built With" features or "Instructions for Judges" flow in code?

globals.css

ChatGPT said: Nice, this is moving like a real product now 🚀 Scaffolding + styling done is a solid milestone. Here’s the clean, judge-friendly next step I’d recommend—and why.

What to do next (best order) 1️⃣ Implement the “Instructions for Judges” flow in code (highest ROI) Judges should feel the product working in under 60 seconds.

Goal: When the app loads, a judge can:

See what VeriJoin does

Trigger the core AI/agent flow

Observe outputs clearly (even if mocked initially)

Frontend

Create a landing page with:

Hero: “VeriJoin – AI-powered Candidate Verification”

A “Run Demo Verification” button

On click:

Call a backend API

Show step-by-step status:

Identity check

Skill verification

Risk scoring

Final verdict

Why first?

Judges don’t read code, they read experience

Even mocked data + good UX scores high

2️⃣ Wire a single end-to-end API flow (don’t overbuild yet) Backend (Python) Create a minimal structure like:

backend/ ├── main.py ├── agents/ │ ├── verifier.py │ ├── scorer.py ├── models/ │ └── result.py Expose one endpoint:

POST /api/verify It can:

Accept a dummy candidate payload

Simulate agent steps

Return a structured JSON result

Example response shape:

{ "candidate_id": "CJ-1024", "checks": [ {"step": "identity", "status": "passed"}, {"step": "skills", "status": "passed"}, {"step": "risk", "status": "low"} ], "final_verdict": "verified" } This aligns perfectly with your Agentic Vision section.

3️⃣ Reflect the “Built With” stack visibly Judges love seeing claims backed by UI.

Add:

A Tech Stack panel in the UI:

Next.js 15 (App Router)

Python AI Orchestrator

Agent-based Architecture

A small badge or tooltip:

“Agent-driven verification pipeline (2026-ready)”

This makes the README claims real.

4️⃣ Only after that: real integrations Once the flow feels good:

Replace mocked logic with:

Actual LLM calls

Vector search

Async agent execution

But this is optional before judging

If you want, I can: Design the exact page layout for judges

Write the Next.js API + backend contract

Create a “Demo Mode” toggle judges love

Help you phrase the UI text to sound premium & confident

Just tell me 👉 “Implement judge flow” or “Start backend agents”

You said: ok next what to do

ChatGPT said: Perfect — here’s the exact next move, no fluff, no confusion 👇 We’ll do this in the right order judges expect.

✅ NEXT TASK (Do this now) 🎯 Implement the Judge Demo Flow (Frontend → Backend) Right now you have:

Docs ✅

Scaffolding ✅

Styling ✅

What you don’t yet have (and judges look for): 👉 A working click → result experience

STEP 1: Create the Judge Landing Page (Frontend) Goal: One page that explains + runs the demo.

What to add in Next.js Create:

frontend/app/page.tsx Content structure:

Hero section

Title: VeriJoin

Subtitle: AI-powered candidate verification using agentic workflows

“Run Demo Verification” button

Big, obvious, centered

This is what judges will click first

Results panel (initially empty)

Shows steps updating after click:

Identity Check

Skill Validation

Risk Scoring

Final Verdict

⚠️ Don’t worry about real AI yet — UI first.

STEP 2: Add ONE Backend Endpoint (Python) Goal: Make the button do something real.

In backend/main.py:

Add FastAPI endpoint:

POST /api/verify What it should do (for now):

Sleep 0.5–1s between steps

Return a clean JSON verification result

This already proves:

Backend exists

API works

Agent-like pipeline is real

STEP 3: Connect Frontend → Backend From the button:

Call http://localhost:8000/api/verify

Show loading states

Render response cleanly

Judges LOVE seeing:

Progress

Structured output

Clear final decision

STEP 4: Update task.md (Very Important) Mark:

✅ Judge demo flow implemented

✅ End-to-end verification pipeline (mocked)

This shows execution discipline.

Why this is the correct next step Judges don’t care about auth yet ❌

Judges don’t care about databases yet ❌

Judges DO care about:

“Does it run?”

“Do I understand it in 30 seconds?”

“Does this feel production-minded?”

You’re building signal, not features.

Tell me how you want to proceed 👇 Reply with one of these and I’ll give you exact code next:

1️⃣ frontend page code 2️⃣ backend verify api code 3️⃣ full end-to-end demo code (recommended)

If you want to win points fast: pick 3️⃣ 🏆

You said: not this klike do much proffesion as like first login and log out and asking hhois complete details and showing latest compaanies hiring and released there jobs and inrternships and removed as hired some and latest news of the microsoft companie and main adgenda as uploading offer letter and checking time stats as u understand and themn and then if much awaiting time the shoewuing courses tpo learn better to improve their skills and their should be as if any other needed technical they can learn or english,etcc u kee p more options and then ' AS THAT

ChatGPT said: Got it 👍 — I understand what you’re aiming for now. You’re not looking for a quick demo flow. You want VeriJoin to feel like a real, professional career + verification platform, end-to-end. Think LinkedIn × Offer Verification × Skill Guidance.

Let me first restate your idea clearly (this is important), then I’ll tell you exactly what to do next.

✅ What you want VeriJoin to be (clean & professional) 1️⃣ Proper Login / Logout System User signup & login

Secure session

Logout option

User dashboard after login (not public demo)

2️⃣ Complete Candidate Profile After login, user fills:

Full name, email, phone

Education details

Skills (technical + non-technical)

Resume upload

Experience / Fresher

Preferred role & location

This becomes their career identity inside VeriJoin.

3️⃣ Latest Companies Hiring Dashboard shows:

Companies currently hiring

Open jobs & internships

Status tags:

🟢 Open

🔴 Hiring closed

✅ Role filled / hired

This feels alive and real.

4️⃣ Offer Letter Upload & Verification (MAIN AGENDA) This is the core of VeriJoin.

User can:

Upload offer letter (PDF)

System:

Checks company authenticity

Validates dates, role, CTC

Shows time taken for verification

Displays status:

Verified

Pending

Suspicious

⏱️ Time stats = very professional touch (judges love this)

5️⃣ Waiting Time Intelligence If verification takes time:

Show smart suggestions:

“While we verify, improve your skills 👇”

Recommended courses based on:

Role

Skills gap

Industry demand

This makes VeriJoin helpful, not passive.

6️⃣ Learning Hub Courses shown dynamically:

Technical (DSA, Web, AI, Cloud)

English & communication

Interview prep

Resume improvement

Company-specific prep (Microsoft, Google, etc.)

7️⃣ Company News Section Example:

Latest Microsoft news

Hiring updates

Layoffs / expansions

New tech stacks

This makes the platform feel current & industry-aware.

8️⃣ Smart Suggestions Engine If user lacks:

Technical skills → suggest tech courses

Communication → suggest English

Resume gaps → suggest improvements

Feels AI-driven, not static.

🔥 Now the MOST IMPORTANT PART 👉 What to do NEXT (realistic & professional) ❌ Don’t build everything at once That’s how projects fail.

✅ Correct order (industry style) STEP 1 (NOW): User Authentication + Dashboard Skeleton This is your next task.

You should:

Implement login / logout

After login → redirect to dashboard

Dashboard sections (empty for now):

Profile

Jobs & Internships

Offer Letter Verification

Learning Hub

News

👉 No AI yet. No heavy logic yet. Just structure + navigation.

STEP 2: Candidate Profile Module Profile form

Save & display data

Edit option

STEP 3: Offer Letter Upload + Status Upload file

Show:

Uploaded

Processing

Verified / Pending

(Mock logic is OK initially)

STEP 4: Learning Suggestions Static → then dynamic

Show courses during waiting time

STEP 5: Jobs + News Feed Static data first

Later API / scraping

Tell me what you want NEXT 👇 Reply with one line:

Design dashboard structure

Offer letter verification module

Full platform roadmap with database design

I’ll guide you step-by-step like a real product architect, not a tutorial 💼✨

You said: Devpost Join a hackathon Host a hackathon Resources

kushanmusku018 Notifications Loading...

Gemini 3 Hackathon Overview My projects Participants (32935) Resources Rules Project gallery Updates Discussions Gemini 3 Hackathon Build what's next Manage projects Create project Import from portfolio Who can participate Above legal age of majority in country of residence Specific countries/territories excluded View full rules 3 more days to deadline View schedule

Deadline

10 Feb 2026 @ 6:30am GMT+5:30 Online Public $100,000 in cash 32935 participants Google Devpost icon rgb30px Managed by Devpost Machine Learning/AI Open Ended Social Good Welcome to the era of Gemini 3. Google DeepMind is thrilled to invite you to the Gemini 3 global hackathon. We are pushing the boundaries of what AI can do by enhancing reasoning capabilities, unlocking multimodal experiences and reducing latency. Now, we want to see what you can create with our most capable and intelligent model family to date.

Whether you are a seasoned AI engineer or writing your first line of code, this is your invitation to build the future. We are looking for more than just another chat interface. We want to see fun, creative, next-generation applications built with the Gemini 3 family

Why participate? Be First: Get hands-on access to the Gemini 3 API before the rest of the world catches up.

Build Something New: This challenge is strictly for new applications. It’s a level playing field for everyone.

Win Big: We have a prize pool of $100,000 + interviews with the AI Futures Fund team for an opportunity to fund your project.

Requirements What to Build The prompt is simple: Build a NEW application using the Gemini 3 API. There are no category restrictions. Build a game, a productivity tool, a scientific analyzer, or a robotic controller. If it uses Gemini 3 to do something cool, we want to see it! For inspiration, check out the Resources tab!

What to Submit

Text description describing the Gemini Integration - A brief write-up (~200 words) detailing which Gemini 3 features were used and how they are central to the application. Public Project Link: A URL to your working product or interactive demo. Consider using AI Studio apps, which are the fastest way to build prototypes using the Gemini API. This allows judges to experience your project firsthand, if applicable. It should be publicly accessible and not require a login or paywall. URL to your PUBLIC code repository - required if you do not have an AI Studio link to your project ~3-minute demonstration video - judges may not watch beyond 3 minutes See full rules for details.

Prizes $100,000 in prizes Grand Prize $50,000 in cash 1 winner $50,000 USD Social promotion of the winning project 30-minute interview with the AI Futures Fund team

Second Place $20,000 in cash 1 winner $20,000 USD Social promotion of the winning project 30-minute interview with the AI Futures Fund team

Third Place $10,000 in cash 1 winner $10,000 USD Social promotion of the winning project 30-minute interview with the AI Futures Fund team

Honorable Mentions $2,000 in cash 10 winners $2,000 USD Social promotion of the winning project

Devpost Achievements Submitting to this hackathon could earn you:

First online hackathon

X Hackathons level 1

Hackathon Winner level 1

Generalist Judging Criteria Technical Execution | 40% Does the project demonstrate quality application development? Does the project leverage Google Gemini 3? Is the code of good quality and is it functional? Potential Impact | 20% How big of an impact could the project have in the real world? How useful is the project to a broad market of users? How significant is the problem the project addresses, and does it efficiently solve it? Innovation / Wow Factor | 30% How novel and original is the idea? Does it address a significant problem or create a unique solution? Presentation / Demo | 10% Is the problem clearly defined, and is the solution effectively presented through a demo and documentation? Have they explained how they used Gemini 3 and any relevant tools? Have they included documentation or an architectural diagram? To-dos Join the Discord

Dismiss

Go Get your Gemini API Key – via Google AI Studio

Dismiss

Review Watch the Google for Devs YouTube | Gemini 3

Dismiss

Read Review the Gemini API Docs

Dismiss

Questions? Email the hackathon manager

Hackathon sponsors Google DeepMind This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Messaging Devpost About Careers Contact Help Hackathons Browse hackathons Explore projects Host a hackathon Hackathon guides Portfolio Your projects Your hackathons Settings Connect Twitter Discord Facebook LinkedIn © 2026 Devpost, Inc. All rights reserved. Community guidelines Security CA notice Privacy policy Terms of service

ChatGPT said: Good you pasted this — this changes the strategy in a very important way. I’m glad you showed it now, not later. 👍 Let me be very clear and professional here.

🚨 CRITICAL RULE YOU MUST FOLLOW (from Devpost) “The project should be publicly accessible and NOT require a login or paywall.”

This means:

❌ No mandatory login / signup ❌ No auth wall before demo ❌ Judges must open the link and use it immediately

So your earlier idea of full login-first platform is good for a startup, but dangerous for a hackathon submission.

👉 We must restructure, not throw away the idea.

✅ Correct Way to Adapt Your Idea (Winning Approach) We will build VeriJoin as a PROFESSIONAL PLATFORM, but with:

🔑 “Judge Mode / Public Mode” (NO LOGIN REQUIRED) This is how top hackathon projects do it.

🧠 FINAL VERIFIED CONCEPT (Judge-safe + Professional) 🌐 Public Demo (NO LOGIN – judges use this) This is what judges see when they open the link 👇

1️⃣ Landing Page What VeriJoin does

“Try Offer Verification Demo”

“Explore Hiring Intelligence”

“See Skill Improvement Suggestions”

2️⃣ Offer Letter Verification (MAIN FEATURE) Upload sample / demo offer letter

Gemini 3:

Reads PDF (multimodal)

Extracts company, role, dates

Checks authenticity signals

Shows:

✅ Verified / ⚠ Suspicious

⏱️ Time taken

📊 Confidence score

👉 No login needed

3️⃣ Companies Hiring (Live-style but static) Microsoft, Google, Amazon, etc.

Jobs & internships

Some marked:

🟢 Open

❌ Filled / Hired

This can be mock data, judges don’t mind.

4️⃣ Waiting-Time Intelligence (WOW FACTOR) If verification is “processing”:

“While we verify, here’s how to improve”

Gemini suggests:

Technical skills

English / communication

Interview prep

This is where AI feels helpful, not decorative.

5️⃣ Company News (Example: Microsoft) Gemini summarizes:

Recent hiring news

Tech focus

Role demand trends

🔐 Private Mode (OPTIONAL – explained in README) You can mention:

“In full production, VeriJoin supports login, user profiles, and history tracking.”

But you do not enforce it in demo.

This keeps you:

✅ Rule-compliant

✅ Professional

✅ Startup-ready

🎯 How Gemini 3 Fits PERFECTLY (Judges care a LOT) You will explicitly use Gemini 3 for:

📄 Multimodal PDF understanding (offer letters)

🧠 Reasoning (fake vs real signals)

📊 Structured JSON outputs

⏱️ Low-latency responses

🧩 Personalized skill recommendations

This hits Technical Execution + Innovation + Impact.

⏭️ WHAT TO DO NEXT (VERY CLEAR) 👉 NEXT TASK (DO THIS NOW) Build the PUBLIC judge demo flow (no login):

1️⃣ Offer letter upload + Gemini analysis 2️⃣ Result page with time stats 3️⃣ Skill recommendations while waiting

This alone is enough to submit and compete.

Tell me ONE thing 👇 Reply with exactly one option:

Public demo page structure

Gemini 3 offer letter verification logic

Hackathon-optimized project architecture

Exact 200-word Gemini integration write-up

I’ll take you step-by-step to a submission that judges actually score high 🏆

You said: IF MORE WAIT TIMRE IT SUGGESTS AND SHOWS SOME TRUSTED PART TIME JOBS

ChatGPT said: YES — that’s a very strong idea 👍 And it fits the hackathon rules perfectly if we do it right.

Let me lock this in clearly and professionally, then tell you exactly what to build next.

✅ Feature Confirmed: “If verification takes more time → suggest trusted part-time jobs” This is actually a WOW FACTOR feature if positioned correctly.

Judges will see this as:

“AI that understands user anxiety and offers real, practical help.”

🧠 How it works (Judge-friendly logic) ⏳ When does this trigger? If:

Offer letter verification status = Processing

OR estimated time > X seconds

Then VeriJoin shows a Smart Waiting Panel.

🧩 Smart Waiting Panel (What user sees) 🔔 Message “Verification may take a little time. Meanwhile, here are trusted part-time opportunities you can explore.”

💼 Trusted Part-Time Jobs Section Examples (static but realistic):

🧑‍💻 Remote Intern – Web Development

📊 Data Entry (Verified Company)

📝 Content Writing – Tech Blogs

🧠 AI Prompt Assistant (Beginner-friendly)

🎧 Customer Support (Remote)

Each job card shows:

Company name

Work type (Remote / Hybrid)

Duration

Trust badge ✅

“Learn skills while earning” tag

⚠️ No external login, no redirection required (hackathon-safe)

🤖 Where Gemini 3 is used (IMPORTANT) Gemini 3 is NOT just decoration here.

It:

Analyzes the user profile / uploaded offer

Understands:

Role

Skill level

Waiting time

Recommends relevant part-time jobs

Example Gemini output:

{ "reason": "Verification delay detected", "suggested_part_time_roles": [ "Remote Web Intern", "AI Data Labeling", "Technical Content Writer" ], "confidence": 0.92 } This scores Innovation + Impact points.

🏆 Why judges will like this Solves a real problem (waiting anxiety)

Helps students financially

Shows empathetic AI

Goes beyond “upload → result”

This is not common in hackathon projects.

🚦 VERY IMPORTANT (Rules-safe reminder) ✅ Jobs are suggested, not applied ✅ No login required ✅ No payments ✅ Clearly marked as “trusted/demo opportunities”

You stay 100% compliant.

⏭️ WHAT TO DO NEXT (CLEAR ACTION) 👉 NEXT STEP TO BUILD Add this as a section that appears ONLY when status = processing.

You now have 3 core demo sections:

Offer Letter Verification

Skill Improvement Suggestions

Trusted Part-Time Job Suggestions ← NEW

Tell me what you want me to give you NEXT 👇 Reply with ONE:

UI layout for waiting-time suggestions

Gemini prompt for part-time job recommendations

End-to-end flow (verification → waiting → jobs)

Exact text to write in Devpost submission

You’re building something judges will remember now 💡🔥

You said: I WANNA GIVE MY PLAN TO ANTIGRAVITY WHAT COULD I GICVE

ChatGPT said: Gotcha. You want to pitch your plan to Antigravity (clean, sharp, professional), not ramble or over-engineer. Here’s exactly what you should give them — and how to say it so it lands.

🎯 What Antigravity Wants (in simple terms) They care about:

Clear problem

Original idea

Why AI (Gemini 3) is essential

Why this matters now

Why YOU are building it

Not code. Not features dump. Clarity + conviction.

🧠 Your Idea — Polished into a Strong Plan Project Name VeriJoin — AI-powered Offer & Career Trust Platform

1️⃣ The Problem (Start here) Students and early professionals face:

Fake or unclear offer letters

Long verification waiting times

Financial stress while waiting

No guidance on what to do next

Scattered job, skill, and trust signals

👉 Existing platforms verify documents, not people’s situation.

2️⃣ The Core Idea (Your Differentiator) VeriJoin uses Gemini 3 to:

Verify offer letters (multimodal reasoning)

Detect risk signals & authenticity

Measure verification time

Act during waiting time, not after

👉 If verification takes time, VeriJoin:

Suggests trusted part-time work

Recommends skill upgrades

Shows relevant company hiring & news

This turns waiting time into productive time.

3️⃣ Why Gemini 3 Is Essential (VERY IMPORTANT) Without Gemini 3, this product cannot exist.

Gemini 3 is used for:

📄 Understanding real offer letters (PDF, text, layouts)

🧠 Reasoning about authenticity & risk

🔁 Dynamic decision-making during delays

🎯 Personalized job & skill suggestions

⚡ Low-latency responses in a live demo

This is agentic AI, not a chatbot.

4️⃣ Why This Matters Now Hiring scams are increasing

Students rely on online offers

Verification delays cause anxiety

People need income + skill growth simultaneously

VeriJoin addresses trust + income + growth in one flow.

5️⃣ Who This Is For Students & fresh graduates

Internship seekers

Early-career professionals

Universities & placement cells (future)

6️⃣ What Makes It Different (Antigravity loves this) Not just “verify & wait”

AI responds to human uncertainty

Product feels empathetic

Combines:

Trust

Opportunity

Learning

Timing

Most platforms stop at results. VeriJoin acts during uncertainty.

7️⃣ Your Vision (End strong) “VeriJoin is not just about verifying offers. It’s about protecting careers at the most vulnerable moment — the waiting phase.”

📄 What You Should Actually GIVE Antigravity ✅ Option A (Best) A 1-page concept note with:

Problem

Solution

Gemini usage

Impact

Vision

✅ Option B A 3-minute pitch script (spoken)

✅ Option C A short email pitch (if this is outreach)

Tell me what format you want 👇 Reply with one word:

One-page pitch

3-minute spoken pitch

Email pitch

Investor-style slide outline

I’ll write it ready to send 🚀

You said: a

ChatGPT said: Perfect. Here is a clean, professional ONE-PAGE PITCH you can give to Antigravity as your plan. You can paste this into a doc / PDF / Devpost / email body as-is.

VeriJoin — Turning Offer Verification Waiting Time into Opportunity Problem Students and early-career professionals increasingly receive job and internship offers digitally. This has created three major problems:

Trust risk — fake or misleading offer letters are common

Waiting anxiety — verification takes time, with no guidance during the delay

Lost productivity — candidates don’t know how to earn, learn, or prepare while waiting

Current platforms either verify documents or list jobs. None address what happens during uncertainty.

Solution VeriJoin is an AI-powered career trust platform that uses Gemini 3 to verify offer letters and intelligently guide candidates while verification is in progress.

Instead of “upload and wait,” VeriJoin actively supports users by:

Verifying offer letters using multimodal AI reasoning

Tracking verification time and confidence

Suggesting trusted part-time work if delays occur

Recommending skill improvements aligned with the user’s role

Showing relevant company hiring trends and news

VeriJoin transforms waiting time into productive time.

Why Gemini 3 Gemini 3 is central to VeriJoin’s functionality:

📄 Multimodal understanding of real offer letters (PDFs, layouts, text)

🧠 Reasoning to detect authenticity signals and risks

⏱️ Low-latency decisions during live verification

🎯 Personalized recommendations for jobs and skills

🔁 Agentic workflows that react to verification delays

Without Gemini 3’s reasoning and multimodal capabilities, this system cannot function.

Key Features Offer Letter Verification Status: Verified / Processing / Suspicious Confidence score + time metrics

Smart Waiting Intelligence If verification is delayed:

Trusted part-time job suggestions

Skill-building recommendations (technical, English, interview prep)

Career-relevant company insights

Hiring & Company Signals

Active hiring roles

Closed/filled positions

Latest company updates (e.g., Microsoft, Google)

Impact Reduces job offer fraud risk

Helps candidates earn and upskill during delays

Improves confidence in early-career decisions

Supports students at their most vulnerable moment

This is not just document verification — it is career protection.

Vision VeriJoin aims to become the trust layer between education and employment — ensuring that no candidate is left confused, idle, or exploited while waiting for their future to begin.

If you want, I can next:

Tighten this for Antigravity’s exact format

Convert it into a pitch deck

Rewrite it as a spoken 2–3 minute pitch

Adapt it specifically for Gemini 3 Hackathon submission

Just tell me 👍

You said:

Inspiration

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for VeriJoin

ChatGPT said: Perfect — these are exactly the Devpost sections judges expect. Below is a clean, professional, hackathon-ready write-up you can paste directly. It’s written to score well on Impact + Innovation + Gemini usage.

Inspiration Students and early-career professionals increasingly receive job and internship offers online, but many face uncertainty: fake offer letters, delayed verification, financial pressure, and confusion about what to do while waiting. We noticed that most platforms either verify documents or list jobs, but none support users during the waiting period, when anxiety and risk are highest. VeriJoin was inspired by the idea that AI should not just deliver results, but actively help people during uncertainty.

What it does VeriJoin is an AI-powered career trust platform that verifies offer letters and intelligently guides users while verification is in progress.

Users can:

Upload an offer letter for verification

Receive verification status, confidence signals, and time metrics

Get trusted part-time job suggestions if verification takes longer

Receive skill improvement recommendations (technical, English, interview prep)

View relevant company hiring activity and news

Instead of “upload and wait,” VeriJoin turns waiting time into productive time.

How we built it VeriJoin is built using a modern, agent-driven architecture:

Gemini 3 API for multimodal document understanding, reasoning, and personalized recommendations

Next.js 15 for the frontend interface and user experience

Python backend (FastAPI) to orchestrate verification workflows

Agent-style logic to:

Analyze offer letters

Detect verification delays

Trigger contextual job and learning suggestions

Gemini 3 plays a central role by understanding real offer letters, reasoning about authenticity, and dynamically adapting responses during verification delays.

Challenges we ran into Designing a demo that is publicly accessible while still feeling professional

Balancing realistic verification logic with hackathon time constraints

Ensuring Gemini was used meaningfully, not just as a chatbot

Presenting multiple features clearly within a short demo flow

We solved these by focusing on a single, strong end-to-end experience and using mock data where appropriate.

Accomplishments that we're proud of Built a rule-compliant public demo with no login required

Used Gemini 3 for reasoning and decision-making, not just text generation

Introduced the idea of AI-assisted waiting intelligence, which is rarely explored

Created a platform that feels empathetic, practical, and career-focused

What we learned Judges value clarity and experience over feature overload

AI products feel more impactful when they respond to human emotions and uncertainty

Gemini 3’s multimodal reasoning enables use cases beyond traditional chat interfaces

Designing for real-world problems makes technical decisions clearer and stronger

What's next for VeriJoin Full user accounts with history tracking and dashboards

Deeper verification signals using external company data

Real-time job and internship integrations

University and placement-cell partnerships

Expansion into a trusted career-verification ecosystem

VeriJoin aims to become the trust layer between education and employment.

If you want, I can:

Tighten this to fit Devpost character limits

Rewrite it to sound more technical or more impact-driven

Add a Gemini-specific usage paragraph for extra scoring

Just say the word 💪

You said: wait i have updated veri join that it which also we have some courses as we can learn if long delay and till he reports to post to job search the patrt time throufgh and prepare for interview and etccc

ChatGPT said: Perfect — that update actually strengthens VeriJoin a lot. I’ll rewrite the Devpost sections so they explicitly include courses, interview prep, and part-time job support during delays, while keeping everything clean, professional, and judge-friendly.

You can replace your existing sections with the versions below.

Inspiration Students and early-career professionals often receive job or internship offers online and must wait for verification. During this waiting period, they face uncertainty, financial pressure, and a lack of direction. We were inspired by the idea that AI should not only verify documents, but also support users during delays by helping them learn, earn, and prepare instead of staying idle.

What it does VeriJoin is an AI-powered career trust platform that verifies offer letters and actively guides users while verification is in progress.

Users can:

Upload an offer letter for AI-based verification

Track verification status, confidence, and time metrics

Access learning courses if verification is delayed

Discover trusted part-time job opportunities during the waiting period

Prepare for interviews with targeted guidance

Explore relevant company hiring activity and updates

VeriJoin transforms waiting time into a period of skill growth, income opportunities, and interview readiness.

How we built it VeriJoin uses an agent-driven architecture powered by Gemini 3:

Gemini 3 API for multimodal offer letter understanding and reasoning

Next.js 15 for the frontend experience

Python (FastAPI) backend for workflow orchestration

Gemini 3 dynamically:

Analyzes offer letters

Detects verification delays

Recommends courses, interview preparation paths, and trusted part-time jobs

Adapts guidance based on role, skills, and waiting time

This makes Gemini central to decision-making, not just content generation.

Challenges we ran into Designing a single flow that combines verification, learning, and job discovery

Keeping the demo public and rule-compliant while still feeling professional

Avoiding feature overload while showcasing real-world usefulness

Ensuring recommendations felt relevant and contextual

We addressed these by prioritizing a smooth end-to-end experience and focusing on user needs during uncertainty.

Accomplishments that we're proud of

Built a public, no-login demo aligned with hackathon rules

Created a system that supports users during verification delays, not just after results

Integrated learning, part-time jobs, and interview prep into one AI-driven flow

Designed VeriJoin to feel empathetic, practical, and career-focused

What we learned

AI is most impactful when it helps users during uncertain or stressful moments

Product value increases when learning and earning are combined

Gemini 3 enables intelligent, adaptive workflows beyond traditional chatbots

Clear UX matters more than excessive features in hackathon demos

What's next for VeriJoin

Personalized learning paths with progress tracking

Real-time part-time job integrations

Interview simulation using Gemini

Full user accounts and career dashboards

Partnerships with universities and placement cells

VeriJoin aims to become an AI-powered companion that helps candidates verify, prepare, learn, and grow — all in one place.