Applied AI
RAG, agents, embeddings and practical LLM workflows
Turning ideas into scalable software through full-stack engineering, applied AI, and thoughtful systems. I solve problems, ship fast, and keep learning every day.
01 Selected work
A selection of AI experiments, privacy-first tools, and developer utilities built to turn complicated workflows into clear experiences.
A local, private articulation coach helping users improve spoken English fluency, confidence, and spontaneity.
A local conversational RAG system built from first principles to explore production-grade retrieval workflows.
An AI-powered workflow for extracting job information and generating targeted cold emails.
A React PDF generator with an inbuilt editor for creating polished, printable documents.
A private browser-based CSV and XLSX viewer that processes files locally in a Dedicated Web Worker without uploading them.
Veilmark is a local-first privacy redactor for images and PDFs. It allows users to redact sensitive information from documents and images without uploading them to a server, ensuring privacy and security.
AdminLens turns administrative documents into plain-English, evidence-backed action plans. It extracts actions, deadlines, payments, warnings, contacts, clarification questions, and an optional reply draft.
Clarity at every layer. An on-demand network and API connectivity workbench built with Next.js. Diagnose DNS, TLS, HTTP, CORS, response headers, and IP connectivity in one evidence-backed report.
Local-first production log analysis for files that are too large or too sensitive for ordinary tools. It extracts errors, warnings, and other insights from logs and presents them in a structured, searchable format.
A calm, context-aware communication workspace that turns rough workplace text into ready-to-send writing for seven destinations: Email, Slack / Teams, Jira Ticket, RCA / Incident, GitHub Issue, Client Update, and Plain Professional.
A focused, browser-based classroom application.
A storybook-style 3D browser game.
A soft, storybook 3D world for little explorers.
02 What I do
Turning ideas into scalable software since October 2020. Skilled and familiar with JavaScript, Node, React, React Native, Angular, Golang, Python, Bash, Linux, SQL, MySQL, PostgreSQL, MongoDB, LangChain, LangGraph, vector databases, Docker, AWS, GCP, Kubernetes, and modern AI tools. I solve problems, ship fast, and keep learning every day, with a passion for creating impact through clean, thoughtful engineering.
RAG, agents, embeddings and practical LLM workflows
Interfaces and APIs designed as one coherent system
Containers, CI/CD and production-ready infrastructure
Relational, document and vector data stores
03 Experience
04 Toolkit
Python / RAG / LLM Applications / Data Engineering
Node.js / Express / NestJS / Go / LoopBack 4
React / Next.js / React Native / Angular / Tailwind CSS
React Native
AWS / GCP / Docker / Kubernetes / AWS Lambda / Firebase
MongoDB / PostgreSQL / MySQL / SQL / Vector Databases / Microservices
LangChain / LangGraph / RAG / LLM Applications / AI Tooling
JavaScript / Golang / Python / Bash / Linux
05 Get in touch
Have a product idea, a difficult engineering problem, or an ambitious AI workflow? I'd like to hear about it.
aliimranadil2@gmail.com