Shashank S.

Senior full-stack developer (8+ years) and Anthropic-certified AI engineer (Claude Code, MCP). I build production SaaS and mobile apps on React, Next.js, Node.js, TypeScript, Python/FastAPI, Supabase, and React Native — with AI agents, custom MCP servers, RAG, and voice AI built in where they add value. Top Rated Plus, 100% Job Success, 69 completed jobs. I ship products people pay for: a B2B lead-gen platform with 1,000+ paying users, a multi-tenant voice AI SaaS, and healthcare apps live on the App Store. WHAT I BUILD Full-Stack SaaS Development — React, Next.js, TypeScript, Node.js, Python, FastAPI, PostgreSQL, Supabase (Auth, RLS, Edge Functions), Redis, Docker, AWS. Multi-tenant architecture, RBAC, Stripe payments, CI/CD. AI Agents & LLM Integration — production Claude, OpenAI (GPT/ChatGPT), and Gemini pipelines: agent orchestration, structured outputs, prompt engineering, RAG (pgvector, embeddings), evals. Anthropic-certified: Claude Code in Action + MCP Advanced Topics. MCP Server Development — custom Model Context Protocol servers connecting Claude to your APIs, databases, and internal tools. I've shipped a production OAuth 2.1 MCP server with per-tool scopes that exposes a live SaaS to Claude — few freelancers have done this in production. AI Code Rescue & Audits — inherited a codebase from Lovable, Replit, Cursor, Bolt, or another developer? I audit and productionize AI-generated code: security review, Supabase RLS policies, architecture cleanup, severity-scored report. 2 years of Replit work across 15+ Upwork projects; 20+ production sites shipped via Lovable. AI Automation & Workflows — n8n, Make, Zapier (I maintain a native Zapier integration my paying users run), webhook architectures, CRM integration (HubSpot, Salesforce, Pipedrive, GoHighLevel, Attio), email and outreach automation with retry and dead-letter handling. Voice AI Agents — I built SweepTalk (sweeptalk.in), a multi-tenant AI voice agent SaaS: real-time calls on LiveKit, ElevenLabs and Deepgram STT/TTS, PSTN/Twilio telephony, 273 tests in CI. AI receptionists, missed-call recovery, outbound calling agents, appointment booking bots. Mobile Apps (iOS + Android) — React Native, Expo, Flutter. Solo developer of CrisisOR (clinical simulation for anesthesiologists, CME-linked) and lead developer of Ask First (AI health app, 5.0★), both live on the App Store; also Vatisha (AI plant care with vision LLM), Luzo, Joyce Meyer Ministries, and GlamourBook. PROOF, NOT PROMISES • Leadscraper — my.leadscraper.de — 1,000+ paying users, AI lead-gen SaaS • SweepTalk — sweeptalk.in — my multi-tenant voice AI SaaS • Altius — altius.trade — production DEX (financial-grade correctness) • CrisisOR, Ask First, Vatisha — live on the App Store HOW I WORK Agent-first with Claude Code daily — AI speed with engineer-grade review. Everything ships with tests, docs, and clean handoff. I flag risks before building, not after delivery. If your last developer (or AI tool) left you something half-finished, my first deliverable is a written map of what actually exists — before I change a line. I respond fast and I'll tell you honestly if I'm not the right fit.