class SauravSaini:
    role       = "Senior Full-Stack & Infrastructure Engineer"
    location   = "India 🇮🇳  |  Available Worldwide 🌍"
    experience = "6+ Years | 20+ Production Projects | 10+ Infra Deployments"
    portfolio  = "https://sauravsaini.dev"
    email      = "contact@sauravsaini.dev"

    expertise = {
        "Backend"    : ["Ruby on Rails 6/7", "FastAPI", "Node.js", "Go", "GraphQL"],
        "DevOps"     : ["Docker", "Kubernetes K8s/K3s", "AWS ECS/EKS/Lambda", "ArgoCD"],
        "AI & ML"    : ["OpenAI GPT-4", "Google Gemini", "LangChain", "RAG", "Embeddings"],
        "Mobile"     : ["React Native", "Flutter", "Offline-First", "App Store Delivery"],
        "Frontend"   : ["Next.js", "React", "TypeScript", "Vue.js"],
        "Databases"  : ["PostgreSQL", "MySQL", "MongoDB", "Redis", "ElasticSearch"],
    }

    currently_building = [
        "RecruitEase Pro → AI multi-tenant hiring platform (FastAPI + Kubernetes)",
        "PanaceaFinder   → AI wellness discovery platform (Gemini + LangChain, SG)",
        "ReScrape        → Realtime B2B scrap marketplace (Rails + ActionCable)",
        "MarketDataHub   → Financial market data infrastructure (FastAPI)",
    ]

    def say_hi(self):
        print("Let's build something amazing together! 🚀")

Full case studies, screenshots & demos → sauravsaini.dev/projects

Stack: FastAPI Next.js Celery Redis MySQL Kubernetes OpenAI WebSockets

Enterprise-grade multi-tenant hiring platform powering large-scale recruitment workflows. Built scalable async Celery AI pipelines for resume parsing, automated scoring, AI proctoring, and technical interview environments — all without blocking user-facing systems. Deployed on production Kubernetes.

AI Proctoring Coding Assessments Resume Parser Multi-Tenant WebSocket Dashboards

Stack: FastAPI Gemini AI LangChain PostgreSQL Next.js Docker

AI-powered wellness & beauty provider discovery. Search by health condition, service type, or location. Built Gemini AI + LangChain semantic search, automated PDF extraction pipelines, and wellness business scraping infrastructure for real-time data enrichment.

Semantic AI Search PDF Extraction LLM Orchestration Location-Aware Discovery

Stack: Node.js FastAPI OpenAI GPT-4 FiasDB Redis Docker

Enables agencies to create custom AI property assistants. Upload datasets → configure GPT-4 chatbot → embed on website. Users discover properties via natural language: budget, bedrooms, location — all answered conversationally via vector semantic search.

RAG Architecture Vector Embeddings CRM Pipeline Realtime Lead Tracking Chatbot SDK

Stack: Go Vue.js WebSockets SQLite JWT   |   License: MIT

Free, self-hosted Docker log viewer with real-time WebSocket tailing, granular RBAC, container lifecycle control, CPU/memory metrics, webhook alerts (Slack/Teams/Discord), and Android + desktop companion apps. Runs at ~30MB RAM — no Elasticsearch required.

Stack: Ruby on Rails React PostgreSQL ElasticSearch Sidekiq OpenAI Vision

Restaurant discovery platform focused on corkage policies & wine intelligence. AI Vision pipeline auto-extracts wine metadata from bottle images via OCR. ElasticSearch powers location-aware restaurant search. Sidekiq handles automated menu scraping & content synchronization.

AI Vision OCR ElasticSearch Sidekiq Async Geolocation Menu Scraping

Stack: React Native TypeScript Redux Toolkit SQLite Stripe Google Maps

Production app used by moving crews across the US. Offline-first SQLite sync for field crews in low-connectivity environments. GPS tracking, digital bills of lading, customer signatures, and Stripe on-site payment collection.

Full details → sauravsaini.dev/services