Ver categorias
Explorar
Fiverr Pro
Português
$
USD
Adi
@marzukrahmanadi
Full Stack Developer
Bangladesh
Bengali, Francês
Sobre mim
Hello, I am Adi, a Full-Stack developer, who has been working on projects as both an individual contributor and as part of a team at various companies for the past several years. More specifically, my experience has been primarily focused on developing web applications that are scalable, smooth and packed with beautiful user experience.
Habilidades
Adi
offline •
Conheça meus serviços
Aplicações Web Full Stack
I will make a webapp with reactjs, nextjs, nodejs and django
Sites Personalizados
I will make a ecommerce website with laravel, wordpress, nodejs
Experiência profissional
Full Stack Developer
CVShelf.com • Autônomo
Dec 2023 - Present • 2 yrs 5 mos
I built CVShelf.com, an AI-powered recruitment platform with bulk CV screening, smart job description generation, and automated candidate outreach using Next.js. RabbitMQ workers handle the heavy lifting: a queue mimics browser request patterns to pull resumes from BDJobs and other platforms, then distributes them in parallel across background workers for sorting, screening, and ranking.
Full Stack Developer
WISM CORP LTD • Período integral
May 2023 - Jan 2026 • 2 yrs 8 mos
I built and maintained three products at WISM, all centered around LinkedIn data and sales automation: Scrupp.com - A LinkedIn and Sales Navigator scraper that lets users extract targeted leads and emails, with Apollo integration for enrichment. I designed the distributed scraping architecture: automated VPS creation via Terraform, Dante proxy setup, and request pattern mimicking to bypass detection — we built our own proxy infrastructure because most proxy services don't support SMTP ports, which we needed for email verification. SalesApi.co - A LinkedIn API for sales and prospecting. Instant lead and email extraction from LinkedIn and Sales Navigator with data enrichment. Sequi.io — A LinkedIn automation tool for building outreach sequences, automating invitations, and generating AI-powered comments, actions to boost engagement. On the architecture side: I used RabbitMQ for microservice communication, distributing messages across server instances with routing keys, and Symfony Messenger for simpler in-app async tasks. We limited concurrent message consumption to 5 per user using Rascal, and monitored queue depth, consumer lag, memory, and connection health via Prometheus with alerts to auto-scale workers before things melted down. The stack is Symfony PHP for the main app serving Vue.js, Python for LLM and embedding tasks, and Hono.js where we needed lightweight APIs — each language chosen for what it's actually good at. I managed the PostgreSQL databases on Scrupp handling millions of rows with optimized queries, and handle all cloud deployments with Docker and Terraform.
