🤖 Interview Bot

An AI-powered interview assistant that represents Sanath Kumar J S in technical interviews using a personal knowledge base, multi-signal RAG search, real-time voice interaction, and a live job market skill gap analyzer.

Built with Next.js 14 · .NET 8 · PostgreSQL + pgvector · Groq LLM · HuggingFace Embeddings · Groq Whisper STT

✨ Features

🧠 3-signal RAG search — Body embedding + title embedding + AI-generated question variants per chunk
🎙️ Voice input — Interviewer asks questions via microphone (Groq Whisper STT)
🔊 Voice playback — Answers read aloud via browser TTS (Web Speech API)
📝 Unanswered question tracking — Questions outside KB stored for prep
📚 Prep dashboard — Review, answer, and promote unanswered questions to KB
📖 Prepare for interview — Read all KB Q&A in a clean study mode with AI-generated interview angles per section
📋 Session history — Browse past interviews with full transcripts and per-message feedback
📊 Confidence scoring — Every answer shows confidence % from weighted vector similarity
🔒 Security — Prompt injection detection, system prompt protection, persona guard
🎯 Skill Gap Analyzer — Live job search from Adzuna + Remotive, skill matching, salary insights, company rankings, auto-digest toggle
🎨 5 chat themes — Obsidian (default), Monolith, Counsel, Signal (light mode), Slate Editorial — selectable from homepage
🛡️ Fallback database — Automatic failover from Supabase to Neon when primary DB is unreachable
📱 Mobile responsive — Works on all screen sizes

🏗️ Architecture

┌─────────────────────┐     HTTP      ┌──────────────────────┐
│   Next.js 14 UI     │ ──────────── │   .NET 8 Web API     │
│   (Vercel)          │              │   (Render.com)        │
└─────────────────────┘              └──────────┬───────────┘
                                                │
                    ┌───────────────────────────┼──────────────────────┐
                    │                           │                      │
          ┌─────────▼──────┐         ┌──────────▼─────┐     ┌────────▼──────┐
          │  PostgreSQL 16  │         │   Groq Cloud   │     │  HuggingFace  │
          │  + pgvector    │         │  LLM + STT     │     │  Embeddings   │
          │  Primary:      │         │  (Free tier)   │     │  (Free tier)  │
          │    Supabase    │         └────────────────┘     └───────────────┘
          │  Fallback:     │
          │    Neon         │         ┌────────────────┐     ┌────────────────┐
          └────────────────┘         │  Adzuna API    │     │  Remotive API  │
                                     │  (Job search)  │     │  (Remote jobs) │
                                     └────────────────┘     └────────────────┘

Database Failover

The DatabaseConnectionManager singleton manages automatic failover:

Primary: Supabase (may pause after 7 days of inactivity on free tier)
Fallback: Neon (serverless, never pauses)
On primary failure, switches to fallback for 5 minutes, then retries primary
/ping runs SELECT 1 to keep both Render and the DB alive
/health returns which DB is active: {"status":"healthy","db":"primary (Supabase)"}

📁 Project Structure

interview-bot/
├── interview-bot-ui/                  # Next.js 14 frontend
│   ├── app/
│   │   ├── page.tsx                   # Home page + theme picker
│   │   ├── chat/page.tsx              # Chat interface (theme-aware)
│   │   ├── prep/page.tsx              # Prep dashboard (PIN protected)
│   │   ├── prepare/page.tsx           # KB reading mode for interview prep
│   │   ├── skill-gap/page.tsx         # Skill Gap Analyzer
│   │   └── sessions/
│   │       ├── page.tsx               # Session list
│   │       └── [id]/page.tsx          # Transcript view with feedback
│   ├── components/
│   │   ├── Navbar.tsx                 # Admin nav bar (theme-aware)
│   │   ├── ThemeProvider.tsx          # React context for chat themes
│   │   └── chat/
│   │       ├── InputBar.tsx           # (theme-aware)
│   │       ├── MessageBubble.tsx      # (theme-aware)
│   │       └── TypingIndicator.tsx    # (theme-aware)
│   └── lib/
│       ├── api.ts                     # All fetch wrappers
│       └── themes.ts                  # 5 chat theme definitions
│
├── interview-bot-api/                 # .NET 8 Web API
│   ├── Controllers/
│   │   ├── ChatController.cs
│   │   ├── KnowledgeController.cs     # KB file list + chunk reader
│   │   ├── TranscribeController.cs
│   │   ├── IngestionController.cs
│   │   └── SkillGapController.cs
│   ├── Services/
│   │   ├── ChatService.cs
│   │   ├── KnowledgeSearchService.cs
│   │   ├── IngestionService.cs
│   │   ├── EmbeddingService.cs
│   │   ├── DatabaseConnectionManager.cs  # Primary/fallback DB failover
│   │   ├── ChunkMetadataHelper.cs
│   │   └── SkillGapService.cs
│   ├── Models/
│   │   ├── ChatModels.cs
│   │   ├── KnowledgeChunk.cs
│   │   └── SkillGapModels.cs
│   ├── knowledge-base/                # Personal KB — .md files
│   │   ├── introduction.md
│   │   ├── career-journey.md
│   │   ├── ai-rag.md
│   │   ├── dotnet.md
│   │   ├── dotnet-interview-qa.md
│   │   └── ...
│   └── Dockerfile                     # Used by Render.com for deployment
│
└── docs/
    ├── schema.sql                     # Core DB schema
    ├── skill_gap_migration.sql        # Skill Gap tables migration
    └── RENDER_MIGRATION.md            # Step-by-step Render.com setup guide

🧠 How RAG Works — 3-Signal Search

Question asked
      ↓
HuggingFace BAAI/bge-base-en-v1.5 → embed query (768d)
      ↓
3 parallel SQL queries (all use HNSW index):
  Signal 1: top 15 by body_embedding      <=> queryVec
  Signal 2: top 15 by title_embedding     <=> queryVec
  Signal 3: top 15 by questions_embedding <=> queryVec
      ↓
Merge all candidates by chunk_id in C#
      ↓
Compute weighted score per chunk:
  titleWeight = title_word_count >= 5 ? 0.30 : 0.15
  bodyWeight  = title_word_count >= 5 ? 0.25 : 0.40
  finalScore  = (questionsSim × 0.45)
              + (titleSim     × titleWeight)
              + (bodySim      × bodyWeight)
      ↓
Tag overlap soft boost (+0.05) + file keyword boost (+0.05–0.08)
      ↓
Confidence gate:
  ≥ 0.62 → HIGH   → Answer from KB
  ≥ 0.52 → MEDIUM → Answer from KB
  < 0.52 → LOW    → Store as unanswered question
      ↓
Groq llama-3.3-70b-versatile generates answer
      ↓
Save to chat_messages with confidence score

Why 3 signals?

Signal	Weight	Purpose
`questions_embedding`	0.45	5 AI-generated question variants per chunk at ingest time. Q↔Q matching is the most accurate signal — solves the Q↔A vector space mismatch.
`title_embedding`	0.30 / 0.15	Section heading embedded alone. Adaptive: 5+ word headings get 0.30; short headings get 0.15.
`body_embedding`	0.25 / 0.40	Semantic content of the full chunk. Safety net for paraphrasing. Gets higher weight when title is short.

🎯 Skill Gap Analyzer

Live job market analysis comparing Sanath's profile against current job postings.

User enters role keywords + location
        ↓
POST /api/skill-gap
        ↓
Adzuna API (IN → GB → US fallback) + Remotive (remote) fetched in parallel
        ↓
Groq LLM extracts required / nice-to-have / trending skills from JDs in batches
        ↓
Compare extracted skills against Sanath's KB-derived skill profile
        ↓
Score each job:
  ATS Score   = keyword match % (Sanath's skills vs JD text)
  Match Score = required skill overlap %
        ↓
Return:
  - Ranked job listings with ATS + match scores
  - Matched skills ✅ | Missing skills ❌ | Trending skills 🔥
  - Salary range (min / median / max from Adzuna data)
  - Top hiring companies ranked by job count
        ↓
Jobs persisted to DB (job_listings table)
User can save jobs → tracked in job_applications table
Auto-digest toggle → settings stored in user_settings table

Skill Gap API Endpoints

Method	Path	Description
POST	`/api/skill-gap`	Run full analysis — fetch jobs + gap report
POST	`/api/skill-gap/save-job`	Save / update job application status
GET	`/api/skill-gap/saved-jobs`	List all saved + tracked jobs
GET	`/api/skill-gap/settings`	Get user settings (auto-digest, keywords)
POST	`/api/skill-gap/settings`	Update user settings

🗄️ Database Schema

CREATE EXTENSION IF NOT EXISTS vector;

CREATE TABLE knowledge_chunks (
    id                   SERIAL PRIMARY KEY,
    source_file          TEXT,
    section_title        TEXT,
    chunk_text           TEXT,
    chunk_index          INTEGER,
    embedding            VECTOR(768),
    topic                TEXT,
    tags                 TEXT[],
    hit_count            INTEGER DEFAULT 0,
    last_used_at         TIMESTAMPTZ,
    created_at           TIMESTAMPTZ DEFAULT NOW(),
    title_embedding      VECTOR(768),
    questions_embedding  VECTOR(768),
    questions_text       TEXT[],
    title_word_count     INT
);

CREATE TABLE interview_sessions (
    id               SERIAL PRIMARY KEY,
    session_code     TEXT UNIQUE,
    company_name     TEXT,
    interviewer_name TEXT,
    round_number     INTEGER,
    started_at       TIMESTAMPTZ DEFAULT NOW(),
    ended_at         TIMESTAMPTZ,
    status           TEXT DEFAULT 'active',
    overall_rating   SMALLINT,
    notes            TEXT
);

CREATE TABLE chat_messages (
    id                SERIAL PRIMARY KEY,
    session_id        INTEGER REFERENCES interview_sessions(id),
    sequence_number   INTEGER,
    role              TEXT,
    message_text      TEXT,
    confidence_score  FLOAT,
    answer_source     TEXT,
    fallback_provider TEXT,
    response_time_ms  INTEGER,
    was_helpful       BOOLEAN,
    created_at        TIMESTAMPTZ DEFAULT NOW()
);

CREATE TABLE unanswered_questions (
    id                  SERIAL PRIMARY KEY,
    session_id          INTEGER REFERENCES interview_sessions(id),
    message_id          INTEGER REFERENCES chat_messages(id),
    question_text       TEXT,
    question_embedding  VECTOR(768),
    times_asked         INTEGER DEFAULT 1,
    status              TEXT DEFAULT 'new',
    priority            TEXT DEFAULT 'low',
    sanath_answer       TEXT,
    kb_chunk_id         INTEGER REFERENCES knowledge_chunks(id),
    question_category   TEXT,
    sanath_answered_at  TIMESTAMPTZ,
    updated_at          TIMESTAMPTZ,
    first_asked_at      TIMESTAMPTZ DEFAULT NOW(),
    last_asked_at       TIMESTAMPTZ DEFAULT NOW()
);

CREATE TABLE session_analytics (
    id                   SERIAL PRIMARY KEY,
    session_id           INTEGER REFERENCES interview_sessions(id) UNIQUE,
    total_questions      INTEGER DEFAULT 0,
    answered_from_kb     INTEGER DEFAULT 0,
    unanswered_count     INTEGER DEFAULT 0,
    avg_confidence_score FLOAT,
    top_kb_files         JSONB,
    weak_topic_areas     JSONB,
    duration_minutes     INTEGER
);

-- Skill Gap tables (run docs/skill_gap_migration.sql)
CREATE TABLE job_listings ( ... );
CREATE TABLE job_applications ( ... );
CREATE TABLE user_settings ( ... );

-- HNSW indexes
CREATE INDEX ON knowledge_chunks USING hnsw (embedding           vector_cosine_ops);
CREATE INDEX ON knowledge_chunks USING hnsw (title_embedding     vector_cosine_ops);
CREATE INDEX ON knowledge_chunks USING hnsw (questions_embedding vector_cosine_ops);

📝 Knowledge Base

14 .md files in interview-bot-api/knowledge-base/. answering-guidelines.md excluded from search.

File	Chunks	Topic
`introduction.md`	7	Who I am, career summary, specialization
`career-journey.md`	5	Toyota Tsusho, Capgemini, Euromonitor, Ingenio
`recent-project.md`	11	RAG systems, JWT migration, integrations
`ai-rag.md`	13	3 production RAG pipelines, tech decisions
`dotnet-interview-qa.md`	20	20 Q&A pairs: DI, async/await, GC, EF Core, Polly
`dotnet.md`	6	Years, strongest areas, design patterns
`leadership.md`	4	3 leadership stories + philosophy
`general-hr.md`	9	Strengths, weakness, salary, notice, education
`my-approach.md`	7	System design examples
`arrays-strings.md`	7	DSA: two pointers, sliding window, hashmap
`trees.md`	6	BST, traversal, DFS vs BFS
`dynamic-programming.md`	4	DP patterns and approach
`complexity-cheatsheet.md`	4	Big-O reference
`answering-guidelines.md`	—	EXCLUDED — internal prompt instructions

Total indexed: 101 chunks across 13 files

🚀 Getting Started (Local)

# 1. Clone
git clone https://github.com/sanathjs/interview-bot.git
cd interview-bot

# 2. Backend
cd interview-bot-api
cp appsettings.example.json appsettings.json
dotnet restore && dotnet run          # http://localhost:5267

# 3. Frontend (new terminal)
cd interview-bot-ui
cp .env.example .env.local
npm install && npm run dev            # http://localhost:3000

# 4. Ingest KB
curl -X POST http://localhost:5267/api/ingest -H "X-Admin-Key: your-key"

⚙️ Configuration

Backend — `appsettings.json`

{
  "DATABASE_URL": "Host=localhost;...",
  "FALLBACK_DATABASE_URL": "Host=ep-xxx.neon.tech;Port=5432;Database=neondb;Username=...;Password=...;SSL Mode=Require;Trust Server Certificate=true",
  "ADMIN_INGEST_KEY": "your-secret-key",
  "LlmProvider": "groq",
  "HuggingFace": { "ApiKey": "hf_..." },
  "Groq": {
    "ApiKey": "gsk_...",
    "Model": "llama-3.3-70b-versatile",
    "BaseUrl": "https://api.groq.com/openai/v1"
  },
  "Adzuna": {
    "AppId":  "your_adzuna_app_id",
    "AppKey": "your_adzuna_app_key"
  }
}

Frontend — `.env.local`

NEXT_PUBLIC_API_URL=http://localhost:5267
NEXT_PUBLIC_PREP_PIN=1234

Render Environment Variables

DATABASE_URL             = (Supabase pooler connection string)
FALLBACK_DATABASE_URL    = (Neon connection string — optional but recommended)
ADMIN_INGEST_KEY         = your-secret-key
LlmProvider          = groq
Groq__ApiKey         = gsk_...
Groq__Model          = llama-3.3-70b-versatile
Groq__BaseUrl        = https://api.groq.com/openai/v1
HuggingFace__ApiKey  = hf_...
Adzuna__AppId        = your_adzuna_app_id
Adzuna__AppKey       = your_adzuna_app_key
ASPNETCORE_URLS      = http://+:10000
PORT                 = 10000

Render Docker services use port 10000 by default.

Vercel Environment Variables

NEXT_PUBLIC_API_URL  = https://your-app.onrender.com
NEXT_PUBLIC_PREP_PIN = your-pin

🔑 API Keys Required

Service	Purpose	Get it at	Cost
Groq	LLM chat + Whisper STT	console.groq.com	Free
HuggingFace	BAAI/bge-base-en-v1.5 embeddings	huggingface.co/settings/tokens	Free
Adzuna	Job search API (India + global fallback)	developer.adzuna.com	Free (1000 calls/day)
Remotive	Remote job search	remotive.com/api	Free, no key needed

📡 API Endpoints

Chat & Sessions

Method	Path	Description	Auth
POST	`/api/chat`	RAG chat — answer + confidence + sources + botSequenceNumber	None
GET	`/api/sessions`	List sessions with stats	None
GET	`/api/sessions/{id}/detail`	Full transcript with per-message feedback (was_helpful)	None
PATCH	`/api/sessions/{code}/details`	Update interviewer name + company (UPSERT)	None
POST	`/api/sessions/{code}/end`	End session	None
PATCH	`/api/messages/{seqNum}/feedback`	Save thumbs up/down (by DB sequence_number)	None
GET	`/api/unanswered`	Prep dashboard questions	None
PATCH	`/api/unanswered/{id}/answer`	Save answer	None
POST	`/api/unanswered/{id}/promote`	Add answer to KB	None
DELETE	`/api/unanswered/{id}`	Delete question	None
POST	`/api/transcribe`	Audio → Groq Whisper → text	None
POST	`/api/ingest`	Re-ingest all KB files	X-Admin-Key header
GET	`/ping`	Keep-alive — runs SELECT 1 to prevent Supabase pause	None
GET	`/health`	Diagnostic — shows active DB (primary/fallback)	None

Knowledge Base (Prepare)

Method	Path	Description	Auth
GET	`/api/knowledge/files`	List all KB files with display names + chunk counts	None
GET	`/api/knowledge/files/{file}`	All sections from a file — title, body, question variants	None

Skill Gap

Method	Path	Description	Auth
POST	`/api/skill-gap`	Fetch jobs + run full gap analysis	None
POST	`/api/skill-gap/save-job`	Save / update job application status	None
GET	`/api/skill-gap/saved-jobs`	List all saved + tracked jobs	None
GET	`/api/skill-gap/settings`	Get user settings	None
POST	`/api/skill-gap/settings`	Update user settings	None

🌐 Deployment

Free Stack — $0/month forever

Layer	Platform	Cost	Notes
Frontend	Vercel	Free forever	Auto-deploys on git push
Backend	Render.com	Free forever	Spins down after 15 min idle
Database (primary)	Supabase	Free forever	500MB limit, pauses after 7 days idle
Database (fallback)	Neon	Free forever	Serverless, never pauses, auto-failover
Embeddings	HuggingFace	Free forever
LLM + STT	Groq	Free forever
Job Search	Adzuna + Remotive	Free forever	1000 calls/day
Keep-alive	cron-job.org	Free forever	Pings `/health` every 10 min
Total		$0/month

Cold start: Render free tier spins down after 15 min idle. First request takes ~30s. Set up a cron-job.org ping to https://your-app.onrender.com/health every 10 minutes to prevent this completely. The /health endpoint runs SELECT 1 which also keeps Supabase from pausing.

Production Stack (future)

Layer	Platform	Cost
Frontend	Vercel	Free
Backend	Render Starter	~$7/mo (always on)
Database	Supabase Pro	~$25/mo
Embeddings	OpenAI text-embedding-3-small	~$0.01/mo
LLM	gpt-4o-mini	~$1–3/mo

Deploy Checklist

# 1. Run Skill Gap migration in Supabase SQL Editor
#    Copy contents of docs/skill_gap_migration.sql → run in Supabase

# 2. Set up Render (see docs/RENDER_MIGRATION.md for full steps)

# 3. Push code — Render + Vercel auto-deploy on push
git add .
git commit -m "your message"
git push origin main

# 4. Re-ingest KB after any .md changes
curl -X POST https://your-app.onrender.com/api/ingest \
  -H "X-Admin-Key: your-key"

# 5. Update NEXT_PUBLIC_API_URL in Vercel → Settings → Environment Variables
#    Set to your Render URL

🛠️ Development Notes

Database & Infrastructure

DatabaseConnectionManager (singleton) wraps all DB access with automatic primary→fallback switching and 5-min retry cooldown
FALLBACK_DATABASE_URL is optional — if not set, app works with primary only (no failover)
/ping runs SELECT 1 via DatabaseConnectionManager.ProbeAsync() — keeps both Render and Supabase alive
/health returns JSON with status + db (which database is active)
UpdateSessionDetailsAsync uses UPSERT — creates session row if PATCH arrives before first chat message
session_analytics has no auto-trigger — stats fall back to live subqueries from chat_messages

Chat & Feedback

ChatResponse includes botSequenceNumber — the DB sequence_number of the bot's reply
Frontend passes sequenceNumber (not React msg-N IDs) to PATCH /api/messages/{seqNum}/feedback
GetTranscriptByIdAsync returns was_helpful per message — shows feedback in session transcript
Prompt injection blocked pre-LLM via IsPromptInjection() — 24 phrases detected

Themes

5 chat themes defined in lib/themes.ts: Obsidian, Monolith, Counsel, Signal (light), Slate Editorial
ThemeProvider (React context) wraps the app; theme saved in localStorage as ib_theme
All chat components use useTheme() hook — const C = useTheme() replaces hardcoded color objects
Theme picker accessible from the palette icon (top-right of homepage)

Knowledge Base

answering-guidelines.md excluded from search at ingest time — do not rename it
Re-ingest required after any KB change — ~5 min for 101 chunks
KB headings should be written as the exact question an interviewer would ask
/api/knowledge/files returns all files with display names (e.g., dotnet-interview-qa.md → "DotNet Interview Q&A")
/api/knowledge/files/{file} strips the "Topic: X\nSection: Y\n\n" prefix from chunk_text for clean reading

Voice & Input

Voice input: MediaRecorder → Groq Whisper → auto-sends transcribed text
Voice playback: window.speechSynthesis (browser TTS, no API needed)

Admin & Auth

Prep dashboard is PIN-protected via NEXT_PUBLIC_PREP_PIN env variable
Admin gear icon visible only when localStorage.ib_role === "admin"
Prepare tab visible in admin nav and admin dashboard homepage

Deployment

Adzuna fetches IN → GB → US; falls back to international if India returns fewer than 10 jobs
job_listings.external_id is unique — re-running analysis updates scores, no duplicates
Render uses port 10000 — set ASPNETCORE_URLS=http://+:10000 and PORT=10000
Keep Render warm: cron-job.org pings /health every 10 minutes

🎨 Chat Themes

5 selectable themes — user picks from the palette icon on the homepage. Choice persists in localStorage.

Theme	Palette	Mode	Vibe
Obsidian (default)	Amber on deep black	Dark	Warm, the original
Monolith	Platinum + Electric Blue	Dark	Cold intellect, zero noise
Counsel	Charcoal + Gold, Serif	Dark	Senior partner's office, gravitas
Signal	Navy + White + Green	Light	Trust-first, corporate-safe
Slate Editorial	Near-Black + Coral	Dark	Editorial confidence, shipped product feel

Themes affect: chat background, message bubbles, avatars, input bar, typing indicator, navbar, follow-up buttons, source banners, bold/list text highlighting, and the live dot color.

🗺️ Roadmap

Phase 2 — Resume & Cover Letter

Upload PDF resume → ingest into KB as resume.md
POST /api/skill-gap/resume → Groq tailors resume to JD → .docx download
POST /api/skill-gap/cover → Groq writes cover letter → .docx download
ATS score with keyword breakdown per job

Phase 3 — Tracking & Automation

Job application status board (saved → applied → interview → offer)
IHostedService daily background job at 9am
Auto-digest — top 10 matched jobs saved daily when enabled

Future

End session UI — star rating + notes modal
Analytics dashboard — KB hit rate, weak topics, confidence trends
ElevenLabs voice cloning (~$6/mo)
Upgrade embeddings to OpenAI text-embedding-3-small

🔐 Security

appsettings.json and .env.local gitignored
Old exposed keys rotated; git history purged via filter-branch
appsettings.example.json and .env.example have no real values
Prep dashboard PIN-protected via env variable
API endpoints have no auth currently — add JWT for production
Prompt injection blocked pre-LLM: 24 phrases in ChatService.cs
System prompt never revealed — LLM-level security rules in Groq system message

📄 License

MIT — feel free to fork and adapt for your own interview assistant.

Built with ❤️ by Sanath Kumar J S

Name		Name	Last commit message	Last commit date
Latest commit History 61 Commits
.claude		.claude
interview-bot-api		interview-bot-api
interview-bot-ui		interview-bot-ui
.gitignore		.gitignore
README.md		README.md
interview-bot.sln		interview-bot.sln

Folders and files

Latest commit

History

Repository files navigation

🤖 Interview Bot

✨ Features

🏗️ Architecture

Database Failover

📁 Project Structure

🧠 How RAG Works — 3-Signal Search

Why 3 signals?

🎯 Skill Gap Analyzer

Skill Gap API Endpoints

🗄️ Database Schema

📝 Knowledge Base

🚀 Getting Started (Local)

⚙️ Configuration

Backend — appsettings.json

Frontend — .env.local

Render Environment Variables

Vercel Environment Variables

🔑 API Keys Required

📡 API Endpoints

Chat & Sessions

Knowledge Base (Prepare)

Skill Gap

🌐 Deployment

Free Stack — $0/month forever

Production Stack (future)

Deploy Checklist

🛠️ Development Notes

Database & Infrastructure

Chat & Feedback

Themes

Knowledge Base

Voice & Input

Admin & Auth

Deployment

🎨 Chat Themes

🗺️ Roadmap

Phase 2 — Resume & Cover Letter

Phase 3 — Tracking & Automation

Future

🔐 Security

📄 License

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Backend — `appsettings.json`

Frontend — `.env.local`

Packages