Zeeshan Ali
Backend engineer building trading systems, internal platforms, and AI-powered products. Currently working with Java, C++, Kafka, PostgreSQL, Python, React, and Next.js.
Current focus
I build backend systems, dashboards, and automation tools for real operational work.
Backend
Trading systems, APIs, and internal tools
AI
Practical AI features and ML experiments
Web
React, Next.js, dashboards, and product UIs
About
I like software that gets quieter as it gets stronger.
I'm a software engineer based in Dubai with experience across trading infrastructure, brokerage operations, defense software, and web products.
I enjoy building systems that remain simple on the surface even when the underlying architecture is complex. Most of my work revolves around backend engineering, distributed systems, internal platforms, and increasingly, AI-assisted products.
C++, FIX-style flows, Kafka, Java services, and PostgreSQL — infrastructure built for real-time execution.
Mission-critical software from naval environments where failure tolerance is near zero.
Full-stack delivery — React, Next.js, backend APIs, cloud deployments, and automation tools.
Real-World Systems
System Architecture
Selected production systems where reliability, latency, and data correctness were critical — trading, streaming, risk, and AI platforms.
These are not demos — each case study reflects real constraints, trade-offs, and failure modes encountered in production.
Selected builds
Things I've Built
Smaller projects, prototypes, and learning builds across web apps, AI/ML, and mobile experiments.
Full stack web Apps
Cyclops AI — Local Coding Agent
A fully local AI coding agent built with LangGraph, FastAPI, and tree-sitter. Seven specialized agents (Profiler, Planner, Navigator, Editor, Verifier, Semantic Verifier, Supervisor) coordinate to plan, navigate, edit, and verify code changes entirely on your machine — no cloud, no API keys. Scored 89/100 on a 10-task real-world benchmark and patched 3/5 real Django bugs from SWE-bench Lite.
Full stack web Apps
Chat Application with Real-time Features
Its a real-time chat application built using React, Node.js, Socket.io, and MongoDB. The application allows users to exchange text and media messages, initiate and answer calls, and manage their friends list.
Full stack web Apps
Task Management App
This full-stack project provides users with a seamless manipulation of managing the famous to do app. Integrated intuitive user interfaces built with React and Next.js on the front end, ensuring smooth navigation and interaction. On the backend, Node.js and Express power the application's functionality, facilitating robust REST APIs for seamless data management. Data persistence is managed through PostgreSQL, ensuring reliability and scalability. Authentication and security measures are implemented using Passport.js, ensuring user data remains protected. Additionally, Watson Backend Logging is utilized for efficient logging and monitoring. The project adheres to industry standards with API testing facilitated through Swagger. The application's UI is enhanced with Tailwind CSS, providing a visually appealing and responsive design.
Full stack web Apps
Stock Market Tracker (Demo)
Exploratory Flutter application demonstrating API integration,state management, and local caching. Built as a learning project,not a production trading system.
Full stack web Apps
Pilot LogBook Applciation
A logbook app for managing Human Resources for Airlines/ Airforce - Currenyly its under development
Full stack web Apps
Scripted saga Website
This website is built with React on the frontend and Flask on the backend. It features custom forms for enhanced user interaction. The backend is powered by Flask, hosted on Google Cloud Compute Engine, with Nginx as the server and Gunicorn managing the application server. The CI/CD pipeline is implemented using GitHub Actions.
Skills
What I work with.
Systems & backend
Data & reliability
AI engineering
Product frontend
Writing
Articles on Medium
Rust, AI/ML, backend systems, and the occasional take on where software is heading.
AI Engineering
1 claps
I Built a Fully Local AI Coding Assistant That Scored 89/100 on Real-World Tasks
No cloud, no API keys, no data leaving your machine. A 7-agent LangGraph pipeline that navigates, edits, and verifies code locally on your GPU.
Read on Medium →AI & Development
92 claps
Beyond the Code: Why Human Judgement Remains Irreplaceable By AI in Software Engineering
Why AI cannot replace software developers — a look at judgement, responsibility, and human review.
Read on Medium →AI Philosophy
236 claps
AI Has Already Taken Over: We Have Surrendered Without Even Reaching AGI
We once painted with words. Now machines think for us. A reflective look at how tools quietly reshape thought.
Read on Medium →Rust Tutorial
10 claps
Let's Learn Rust: Task Manager CLI with Rust Fundamentals Part 2
Building a practical Task Manager CLI application with Rust.
Read on Medium →Rust Tutorial
14 claps
Let's Learn Rust: Task Manager CLI with Rust Fundamentals Part 1
Practice Rust fundamentals in a real CLI project.
Read on Medium →Rust Tutorial
20 claps
Let's Learn Rust 005: Traits and Generics
Understanding Rust's powerful trait system and how generics keep code flexible without sacrificing performance.
Read on Medium →AI Analysis
100 claps
Artificial General Intelligence: Hype, Reality, and the Road Ahead
Is AGI real or just a story we tell ourselves? Cutting through the hype with practical analysis.
Read on Medium →AI/ML
54 claps
How To Build a Sentiment Analysis Model End-to-End Using Python
A full deep learning pipeline from data collection to model deployment.
Read on Medium →Research
Published Papers
Preprints on distributed systems correctness, formal verification, and financial system safety.
Deterministic Execution and Margin Enforcement in Financial Transaction Systems
Empirical Evaluation with Formal Safety Verification
Examines how timing, ordering, and state consistency affect correctness in real trading systems. Evaluates margin enforcement under realistic execution constraints using formal safety verification — grounded in production systems where determinism is a risk control requirement, not a theory problem.
Auth Invariant Tester: Temporal Authorization Correctness Verification for Distributed Systems
Examines failure modes caused by cache staleness, policy propagation delays, and long-lived credentials. Introduces executable invariants such as Policy Epoch Safety to detect them — comparing real authorization implementations against a synchronous reference authorizer to expose correctness bugs that conventional tests typically miss.
Work history
Experience
2023 — Present
Senior Backend Engineer
@ River Prime
Building low-latency trading systems and backend services — C++ bridges, Kafka pipelines, order routing, and risk infrastructure.
2018 — 2023
Software Engineer
@ Naval Forces of Pakistan
Developed mission-critical software for defence operations in environments where failure tolerance is very low.
2016 — Present
Full-Stack Engineer (Freelance)
@ Toptal
Delivered web apps, dashboards, and automation tools for clients across different industries.
Contact
Want to talk about a role, project, or technical problem?
I'm open to backend engineering, systems work, AI product development, and focused consulting.