Overview

Aadil Salman Butt

Full Stack Engineer

Selected workGet in touchProfile

At a glance

Full Stack Engineer

Lahore, Pakistan

Core stack

JavaScriptTypeScriptNode.jsDjangoMachine LearningLLMs
Driver profile

About

Who I am, what I focus on, and how I prefer to work with teams and clients.

Name

Aadil Salman Butt

Role

Full Stack Engineer

Professional summary

I design and build production web and AI systems—full stack from APIs and data to UX—for teams that need reliability, not demos. Experienced with scalable apps, RAG and LLM integrations, MLOps-style pipelines, payments, and performance-minded delivery.

Open to senior full-time roles and selective consulting on product builds and AI integrations.

Stack layers

System architecture

Technologies I use to ship products end to end—from interfaces and APIs to AI, data, and cloud operations. Hover a layer to explore.

Frontend

  • JavaScript
  • TypeScript
  • React.js
  • Next.js
  • HTML5
  • CSS3
  • Responsive Design
  • Shadcn UI

Backend

  • Node.js
  • Django
  • ASP.NET
  • REST APIs
  • Microservices
  • MongoDB
  • MS SQL Server
  • Python
  • C#
  • SQL

AI / ML

  • Machine Learning
  • LLMs
  • RAG
  • FAISS
  • Predictive Modeling
  • Chatbots
  • Voice Assistants

Cloud / Ops

  • Model Deployment
  • Automated Retraining
  • AWS
  • CI/CD
  • Docker
  • Git
  • Linux
  • Jira
  • Agile/Scrum
  • Testing
Selected work

Case studies

A closer look at how I scope problems, make technical choices, and deliver outcomes—written for recruiters and hiring managers as well as clients evaluating product or AI work. Each story maps to a public repository you can review in detail.

ApexFlow

Real-time MLOps for lap-time prediction

AI / MLOps
Context
Motorsport analytics need low-latency predictions and models that stay accurate as telemetry and track conditions change—batch notebooks are not enough on race weekend.
My role
End-to-end system design: data pipeline assumptions, training/retraining hooks, and deployment path suitable for time-sensitive inference.
Execution
Structured the platform around professional-style telemetry, automated retraining loops, and cloud-native orchestration so predictions can be refreshed without manual rewrites. Emphasized reliability boundaries between training, validation, and online inference.
Results
Demonstrates production MLOps thinking beyond a single model file: lifecycle, automation, and operational constraints as first-class concerns—relevant to senior ML/platform roles and to clients shipping predictive products.
PythonMLOpsCloud orchestrationReal-time inference
View repository

LegalMindRAG

RAG over legal corpora with grounded answers

AI / NLP
Context
Legal and policy questions need answers tied to real sources; generic LLM replies risk hallucination and low trust.
My role
Retrieval architecture, chunking and embedding strategy, and LLM prompting so outputs cite the underlying documents (e.g. Universal Declaration of Human Rights and related texts).
Execution
Combined vector search (FAISS) with an LLM layer for retrieval-augmented generation, focusing on source-grounded responses rather than open-ended generation. Tuned the pipeline for domain-appropriate retrieval quality.
Results
Shows how to ship trustworthy AI features where provenance matters—directly transferable to support bots, internal knowledge bases, and compliance-sensitive assistants for freelance or product teams.
PythonFAISSLLMsRAG
View repository

ResearchOS

Multi-tenant SaaS with autonomous research agents

Full Stack / AI
Context
Turning open-ended research objectives into repeatable, professional outputs at scale requires background jobs, isolation between users, and agent orchestration—not a single synchronous API call.
My role
Full-stack product engineering on a modern stack: tenancy-aware workflows, async job processing, and agent graphs that decompose objectives into reports.
Execution
Built on Next.js 14 with Convex for data and LangGraph for agent flows, emphasizing multi-tenant safety and performance so research runs become a managed pipeline rather than ad hoc prompts.
Results
Signals senior-level SaaS and AI integration skills: auth/tenancy patterns, background work, and LLM orchestration—strong for senior full-stack roles and for consulting on AI-native products.
Next.js 14TypeScriptConvexLangGraph
View repository

Churn Predictor

Explainable churn intelligence for retention teams

ML / Analytics
Context
Retention teams need more than accuracy—they need risk scores, drivers of churn, and interfaces that non-ML stakeholders can use.
My role
Modeling with gradient boosting, explainability layer, API surface, and a Streamlit experience for exploration and dashboards.
Execution
Used XGBoost with SHAP for interpretability, exposed predictions via FastAPI, and paired with Streamlit for interactive analysis so business users can act on insights.
Results
End-to-end ML product: from training to API to UI—exactly the kind of delivery clients pay for and hiring managers look for in senior data/ML engineers.
PythonXGBoostSHAPFastAPIStreamlit
View repository

InsightOS

Natural language to safe SQL and CFO-style insights

Full Stack / AI
Context
Analysts and leaders want to ask questions in plain language without risking destructive SQL or leaking data across users.
My role
Designed per-user data isolation, read-only query execution, and NL-to-SQL (or equivalent) flows with background processing for heavier insight generation.
Execution
Users upload or connect schemas; the system constrains queries to safe, read-only operations and generates narrative insights suitable for finance-style reporting—balancing flexibility with guardrails.
Results
Demonstrates judgment on security and product shape for AI + data apps—valuable for enterprise-facing freelance work and senior backend/full-stack positions.
TypeScriptSQLBackground jobsLLM-assisted analytics
View repository
Additional work

Projects

Other public repositories and experiments—short summaries mirror each repo's GitHub description. Flagship work with more context lives in Case studies above.

FairPlayReviewSystem

A comprehensive cricket analysis system for fair play review, featuring advanced ball detection with hybrid tracking, batsman identification and tracking, wicket detection, and pose estimation.

Python
2Repo

FinBotics

Finbotics is a modern finance dashboard application designed to help small businesses visualize their financial data and gain actionable insights through machine learning predictions. The application is built using the MERN stack and features interactive charts, analytics, and predictive modeling to streamline business decision-making.

JavaScript
1Repo

MdcatExpert

MDCAT Expert is an AI-powered, data-driven preparation tool designed to help students excel in the MDCAT exam. It provides personalized study plans, smart question practice, performance analytics, and targeted insights to identify weaknesses and improve efficiently, making exam preparation more strategic and effective.

TypeScript
0Repo

TourAgency

TourAgency is a production-oriented tour booking and operations app for Pakistan adventure travel. Guests can book without an account; customers get a dashboard; staff use a realtime admin panel for tours, unified bookings, users, blog posts, and analytics. An AI trip planner suggests itineraries from your live tour catalog.

PHP
0Repo

Zero Trust Network Segmentation Visualizer

Streamlit app to model a small network, define simple Zero Trust policy rules (allow/deny), and visualize allowed vs risky communications with a simple risk score and explanations.

Python
0Repo

Modbus Traffic Analyser

Parse Modbus/TCP traffic from a **PCAP** (TCP port 502) or from simulated logs, extract protocol fields (function codes, reads/writes, register ranges, unit id, src/dst), and flag suspicious activity via a simple rule engine.

Python
0Repo

OT Network Anomaly Detection Dashboard

Lightweight Streamlit dashboard for OT/industrial network anomaly detection using unsupervised ML (default: Isolation Forest).

Python
0Repo

AI Powered GitHub Repository Secret Leakage Detector

AI-Powered GitHub Repository Secret Leakage Detector is a hybrid system that combines AI models with rule-based checks to automatically detect and prevent accidental exposure of sensitive data in code repositories, helping identify secrets, credentials, and API keys before they are committed or shared.

TypeScript
0Repo

Portfolio Website

High-performance personal portfolio: F1-inspired dark UI, JSON-driven resume, GitHub projects with ISR, and a Groq-powered chat assistant.

TypeScript
0Repo

Hospital Management System

Hospital Management System is a full-featured ASP.NET-based application designed to manage hospital operations end-to-end, including patient records, appointments, doctor management, billing, admissions, and staff workflows. It provides a centralized, secure, and efficient system to streamline administrative tasks, improve data management.

C#
0Repo

Magik

Magik is a lightweight, imaginative programming language built on LLVM that reimagines coding as spellcasting, offering a minimalistic, expressive, and fun way to write programs while retaining modern development capabilities.

C
0Repo

F1 Tyre Strategy Predictor

A data-driven predictive analytics project designed to forecast Formula 1 race tyre strategies and driver performance outcomes. This project combines multiple machine learning models to help race strategists, analysts, and motorsport enthusiasts gain insights into potential race outcomes based on historical race data and grid positions.

Jupyter Notebook
0Repo

MotorAi

MotorAI isn't just another car marketplace—it's an intelligent platform that leverages cutting-edge AI to transform the car buying and selling experience. Built with modern technologies and designed for scale, this project demonstrates enterprise-level full-stack development skills

JavaScript
0Repo

Aadilsal

Config files for my GitHub profile.

configgithub-config
0Repo

DoodleJump

Clone of Doodle Jump using assembly language 8086

Assembly
0Repo

CandyCrush

First semester project of creating a clone of candy crush using basic programming concepts

C++
0Repo

Tetris

Tetris project using the concepts of Object Oriented Programming.

C++
0Repo
Live telemetry

GitHub activity

A live view of public repos I'm actively maintaining—useful for seeing where recent commits and iteration are happening.

Race history

Experience

Roles where I owned delivery across the stack: product-facing features, integrations, AI assistants, payments, and performance.

  1. Associate Software Engineer

    Oct 2025 – Present

    ILMVERSITY (Da1ilmverse)

    • Built full-stack web applications using React, Next.js, and Node.js delivering seamless user experiences
    • Developed chatbots and voice assistants to enhance engagement and automate support
    • Integrated Stripe, Chase Bank, and KuickPay improving payment reliability
    • Optimized frontend performance and backend APIs to improve speed and responsiveness
    • Created reusable UI components and scalable design patterns
    • Collaborated with cross-functional teams to deliver production-grade features

  2. Web Development Intern

    Jul 2023 – Mar 2024

    Fiveriver Technologies

    • Built responsive frontend features using JavaScript and React
    • Developed API-driven applications with MongoDB backend
    • Refactored legacy code and implemented unit testing to reduce technical debt
Credentials

Education & certifications

Academic background and formal credentials alongside hands-on engineering and product delivery.

Education

  • BS Computer Science

    National University of Computer & Emerging Sciences (FAST), Lahore

    Aug 2022 – Present

  • A Levels

    Lahore Grammar School, Johar Town

    Aug 2020 – Jun 2022

  • O Levels

    Beaconhouse School System, Garden Town

    Aug 2017 – Jun 2020

Certifications

  • AWS Academy Cloud Foundations
  • Digital Skills: Artificial Intelligence (FutureLearn)
Performance

Metrics

At-a-glance signals from public GitHub activity—best read together with case studies and experience above.

Total repos

22

Years experience

2+

Public commits

229

Public commits attributed to you in GitHub's search index.

Reach out

Contact

Whether you're hiring for a senior role or need a focused build or integration, email and WhatsApp are the quickest ways to reach me. I respond to both full-time and consulting inquiries.

aadilsalman786@gmail.comhttps://github.com/aadilsalWhatsApp
Send emailMessage on WhatsApp