● Worked as a staff-level senior full-stack engineer on a multi-vendor enterprise mobile application, owning end-to-end delivery across mobile architecture, a FastAPI and Python backend on Azure, iOS enterprise distribution, and cross-vendor compliance. ● Architected state management migration from Context API to Zustand, reducing render cycles on complex data grids. Migrated authentication from MSAL SDK to expo-auth-session with PKCE flow against Azure AD. ● Designed an end-to-end encrypted notes module (AES-256, PBKDF2 key derivation from PIN, and biometric unlock), meeting enterprise compliance requirements without server-side key access. ● Led iOS enterprise distribution combining Apple Developer Enterprise Program (ADEP) and Microsoft Intune MDM. Owned EAS Build CI/CD on Azure Pipelines, set up Detox and Maestro E2E tests, and integrated an AI assistant with a LangChain backend.

● Architected and shipped a React Native field-sales app for a US pharma client, leading mobile and FastAPI/Azure backend across HIPAA-aware infrastructure and iOS distribution. ● Built and solo-shipped a React Native dental practice management app whose parent company was acquired by a NASDAQ-listed healthcare distributor (Henry Schein, HSIC) in 2025. ● Delivered production features for an a16z-backed Web3 wallet infrastructure startup (Dynamic.xyz), shipping Solana and EVM integrations consumed by hundreds of downstream applications. ● Designed and built a RAG classification pipeline using pgvector, OpenAI embeddings, NestJS, Next.js, and Trigger.dev to automatically tag and route inbound business signals at scale. ● Operated a self-hosted production homelab on a dedicated server (Coolify, Docker, Cloudflare Tunnel, Tailscale, Trigger.dev v3) running internal automation, MCP servers, and side-project workloads. ● Built FlowBot, a no-code automation platform with 170+ node types orchestrating Telegram, Discord, and WhatsApp workflows for non-technical operators. ● Engineered a Telegram OSINT crawler combining Neo4j graph storage, PostgreSQL, and the MTProto protocol (mtcute) to map cross-channel relationships across millions of messages. ● Reverse-engineered and modified firmware for a locked-down embedded Android projector, publishing recovery and unlock tooling consumed by a small but active community of device owners. Projects: 1. FlowBot. Technologies & Skills: Python, FastAPI, OpenAI, LangChain, Full-stack, Trigger.dev, User Interface, Mobile UI, GitHub. ● Built a no-code automation platform for Telegram, Discord, and WhatsApp with 172+ node types and 386+ commits. ● Developed a visual workflow editor that enabled users to create complex, multi-step automations across messaging platforms without writing code. ● Implemented features such as flow chaining, variable interpolation, webhook integrations, conditional branching, scheduled triggers, and real-time message processing. ● Built the system as a TypeScript monorepo using Telegram MTProto (mtcute), the Discord Gateway, the WhatsApp Business API, a NestJS backend, a Next.js dashboard, PostgreSQL with Prisma, and Trigger.dev workers. 2. FSC Classifier. Technologies & Skills: Python, OpenAI, Full-stack, Trigger.dev, Responsive Web Design (RWD), Mobile UI, GitHub. ● Built a RAG-based classifier that mapped companies and products to U.S. Federal Supply Class (FSC) codes used in DLA defense procurement. ● Crawled company websites and PDF capability statements using Puppeteer and Cheerio, parsed documents, and ran a hybrid pipeline with pgvector similarity search against previously classified companies as a long-term cache, with GPT-4o reranking and JSON-mode structured output as a fallback. ● Orchestrated asynchronous workflows using Trigger.dev with progress streaming to a Next.js front end. ● Built the system using NestJS, Prisma, PostgreSQL with pgvector, and OpenAI embeddings. 3. Telegram Crawler. Technologies & Skills: Python, FastAPI, PostgreSQL, Full-stack, Trigger.dev, GitHub. ● Built a self-hosted Telegram crawler for OSINT and social graph analysis. ● Used MTProto (mtcute) with a multi-account session pool, Redis Bloom filters for deduplication, and BullMQ for job orchestration. ● Stored users, groups, and messages in a hybrid datastore, using Neo4j as the primary graph database for member relationships, invite link traces, and account clustering via Cypher, and PostgreSQL for operational data and full-text search. ● Built the system as a TypeScript Bun monorepo with Trigger.dev workers and a SvelteKit dashboard.

● Engineered full-stack SaaS products end-to-end for multiple US-based startups across healthcare, Web3, AI, and consumer verticals (DirectDental, Globality, PublicSQ, Willow, and Dynamic.xyz). ● Architected and shipped the DirectDental React Native mobile app from scratch, integrating with an existing Ruby on Rails API and contributing to the company’s 2025 acquisition by Henry Schein (Fortune 500, NASDAQ: HSIC). ● Built and maintained CI/CD pipelines (GitHub Actions and Azure DevOps), cutting release cycles and enabling continuous delivery across web and mobile platforms. ● Prototyped AI and RAG workflows and agentic automation tools using LangChain, OpenAI, and FastAPI, delivered as internal productivity and data-processing solutions.