SynapseForce

AI-Powered Workforce Intelligence Platform

SynapseForce is a full-stack HR management platform that uses skill-based AI matching to automatically assign the right employees to the right projects. HR admins get a complete workforce overview — from resume parsing and skill detection to project tracking, team suggestions, and real-time messaging. Employees get a personal workspace to manage their profile, track assigned projects, and communicate with HR.

---

Features

HR Admin

• Dashboard — workforce overview with active projects, skill gaps, overdue alerts, and a live activity feed
• Employee Directory — searchable table with availability status, top skills, performance score, and project count; expand any row to see full skill breakdown and endorse skills
• Project Management — create projects with required skills; team is auto-assigned via AI skill matching; track progress, deadlines, and status
• Resume Management — upload employee resumes (PDF/DOCX); Apache Tika extracts text and auto-detects skills with strength levels
• Team Builder — enter required skills and get ranked employee suggestions with match scores
• Analytics — skill distribution chart, skill gap analysis, export to CSV, recent activity feed
• Messaging — initiate conversations with any employee; shared inbox with unread counts
• HR Profile — personal profile with workforce stats, top performers, project status breakdown

Employee

• Dashboard — personal stats (skills, projects, performance score), skill strength bars, assigned project progress
• My Profile — view and update availability status (Available / Busy / On Leave), upload resume, see detected skills with endorsement badges, update project progress
• Message HR — direct messaging with the HR team; notifications on new messages

Platform

• JWT authentication with role-based access control (ADMIN / USER)
• Real-time notification bell with unread count badge (polls every 5 seconds)
• Global search across employees and projects
• Dark / light mode toggle
• Fully responsive layout

---

Tech Stack

Backend

Technology	Version	Purpose
Java	17	Language
Spring Boot	3.2.0	Application framework
Spring Security	6.x	Authentication & authorization
JJWT	0.11.5	JWT token generation and validation
Spring Data JPA / Hibernate	6.x	ORM and database access
Apache Tika	2.9.1	Resume text extraction (PDF, DOCX)
PostgreSQL	15+	Production database (Supabase)
H2	—	In-memory database for local development only
Lombok	—	Boilerplate reduction

Frontend

Technology	Version	Purpose
React	18.2	UI framework
Vite	5.1	Build tool and dev server
Tailwind CSS	3.4	Utility-first styling
React Router	6.22	Client-side routing
Axios	1.6	HTTP client
Recharts	2.12	Charts and data visualization
Lucide React	0.344	Icon library

Infrastructure

Service	Purpose
Render	Backend hosting (Docker)
Vercel	Frontend hosting
Supabase	Managed PostgreSQL database (production)
GitHub	Source control and CI/CD trigger

---

Project Structure

SynapseForce/
├── backend/                          # Spring Boot API
│   ├── src/main/java/com/synapseforce/
│   │   ├── auth/                     # Login, register, JWT
│   │   ├── user/                     # User entity, controller, availability
│   │   ├── skill/                    # Skill entity and repository
│   │   ├── resume/                   # Resume upload, Tika parsing, skill extraction
│   │   ├── project/                  # Project CRUD, progress tracking, deadlines
│   │   ├── task/                     # Task management
│   │   ├── message/                  # HR ↔ employee messaging
│   │   ├── notification/             # In-app notifications
│   │   ├── activity/                 # Activity log feed
│   │   ├── analytics/                # Overview stats, skill gap, export
│   │   ├── team/                     # AI skill-matching team suggestions
│   │   ├── security/                 # JWT filter, security config
│   │   └── common/                   # Data seeder, global exception handler
│   ├── src/main/resources/
│   │   ├── application.properties    # Local (H2) config
│   │   └── application-prod.properties # Production (PostgreSQL) config
│   └── Dockerfile                    # Multi-stage Docker build
│
└── frontend/                         # React + Vite SPA
    └── src/
        ├── pages/                    # Dashboard, Employees, Projects, Messages, etc.
        ├── components/
        │   ├── layout/               # Sidebar, Topbar, AppLayout
        │   └── ui/                   # StatCard, EmptyState, SkillBadge
        ├── context/                  # AuthContext, ThemeContext
        └── lib/                      # Axios instance with JWT interceptor

---

Local Development

Prerequisites

• Java 17+
• Maven 3.9+
• Node.js 18+

Backend

cd backend
./mvnw spring-boot:run

The API starts on http://localhost:8081. H2 in-memory database is used automatically for local development — no setup needed. In production, the app connects to Supabase PostgreSQL via the DATABASE_URL environment variable. Demo data (users, skills, projects) is seeded on first run.

H2 console: http://localhost:8081/h2-console

• JDBC URL: jdbc:h2:mem:synapseforce
• Username: sa / Password: (empty)

Frontend

cd frontend
npm install
npm run dev

App runs on http://localhost:5173. The Vite dev server proxies /api requests to localhost:8081.

---

Deployment

Backend — Render

1. Create a new Web Service on Render, connect the GitHub repo
2. Set Language to Docker, Root Directory to backend
3. Add environment variables:

Variable	Value
DATABASE_URL	Your Supabase PostgreSQL connection string
JWT_SECRET	Any long random string
FRONTEND_URL	Your Vercel frontend URL

Render injects a PORT env var automatically — the Dockerfile handles it.

Frontend — Vercel

1. Import the GitHub repo on Vercel
2. Set Root Directory to frontend
3. Add environment variable:

Variable	Value
VITE_API_URL	Your Render backend URL (e.g. https://synapseforce-api.onrender.com)

Database — Supabase

Create a free project on Supabase. Copy the PostgreSQL connection string from Settings → Database → Connection string (URI). Hibernate auto-creates all tables on first boot (ddl-auto=update).

---

API Reference

Auth

Method	Endpoint	Access	Description
POST	/auth/register	Public	Register a new user
POST	/auth/login	Public	Login and receive JWT

Users

Method	Endpoint	Access	Description
GET	/users	Admin	List all employees
GET	/users/me	Any	Get current user profile + skills
GET	/users/{id}	Admin / Self	Get user profile
GET	/users/{id}/skills	Any	Get user skills
PATCH	/users/{id}/availability	Admin / Self	Update availability status
PATCH	/users/skills/{skillId}/endorse	Admin	Toggle skill endorsement

Projects

Method	Endpoint	Access	Description
GET	/projects	Any	List all projects
POST	/projects	Admin	Create project (auto-assigns team)
PATCH	/projects/{id}/status	Admin	Update project status
PATCH	/projects/{id}/progress	Admin / Member	Update progress % and note
DELETE	/projects/{id}	Admin	Delete project
GET	/projects/my/{userId}	Any	Get projects for a user

Resumes

Method	Endpoint	Access	Description
POST	/resume/upload	Any	Upload resume, extract skills
GET	/resume/all	Admin	List all uploaded resumes
GET	/resume/my	Any	Get current user's resumes

Team & Analytics

Method	Endpoint	Access	Description
GET	/team/suggest?skills=Java,React	Admin	Get ranked team suggestions
GET	/analytics/overview	Any	Dashboard stats, skill gaps, activity
GET	/export/employees.csv	Admin	Export employee data as CSV
GET	/export/projects.csv	Admin	Export project data as CSV

Messages & Notifications

Method	Endpoint	Access	Description
POST	/messages/send/{receiverId}	Any	Send a message
GET	/messages/conversations	Any	List conversations
GET	/messages/conversation/{otherId}	Any	Get message thread
GET	/messages/admin	Any	Get HR admin contact
GET	/notifications	Any	Get notifications
GET	/notifications/unread-count	Any	Get unread count
PATCH	/notifications/{id}/read	Any	Mark notification as read
PATCH	/notifications/read-all	Any	Mark all as read

---

How AI Team Matching Works

When a project is created with required skills (e.g. Java, Spring Boot, Docker):

1. The TeamSuggestionService fetches all employees and their skill records
2. Each employee is scored based on how many required skills they have and the strength level of each matching skill
3. Employees are ranked by score and the top matches are assigned to the project
4. Each assigned employee receives an in-app notification

The same scoring logic powers the Team Builder page, where HR can enter any skill set and get a ranked list of best-fit employees before creating a project.

---

Screenshots

---

Developed by

Smriti Prajapati