Madi Turysbek

Backend Developer - SDF

DSRC

Built shared microservice authorization and an automated proposal-routing platform using Kafka-synced data.

• Replaced centralized gateway authorization due to security risks by building a shared auth-client package that gates each microservice behind plain nginx with Kafka-synced user and permission data. • Automated complex proposal routing for a government tax authority by building an end-to-end platform with a multi-step state machine and Kafka-synced org data. • Tech Stack: Python, FastAPI, PostgreSQL, ClickHouse, Docker

AI Engineer - ASK71

AI71

Abu Dhabi-based artificial intelligence company.

Reduced the number of models needed in production assistant deployments and built an MCP-based interactive visualization tool.

• Reduced production AI model footprint by having assistant tools reuse the main conversation model by default, so clients run fewer models on limited GPU and on-prem works end-to-end. • Extended agent capabilities by building an MCP-based data visualization tool that turns agent outputs into interactive visualizations through a standalone containerized frontend. • AI/Tech Stack: Python, FastAPI, FastMCP, LangGraph, RAG, Agentic AI, LLMs, React, Redis

Fullstack Developer - Damumed Clinic

Damumed Business

Built clinic appointment booking and doctor scheduling systems with configurable availability and timing rules.

• Designed and Implemented an appointment reservation system with dynamic availability checks, validation rules, and support for cross-clinic bookings • Built a doctor scheduling system with configurable work types, automated slot generation, and support for complex timing constraints • Tech Stack: C#, ASP.NET, Vue.js, Javascript, Docker, PostgreSQL

Backend Developer - Digital Monitoring & Management System

AG TECH

Raised object positioning and routing accuracy from 70% to 95% using predictive corrections, and added a gRPC intermediary service.

• Implemented positioning correction strategy using predictive algorithms, resulting in a significant increase in accuracy from 70% to 95% for both positioning and routing of objects. • Established intermediary service using gRPC for seamless communication between services and file storage, optimizing data management. • Tech Stack: C#, ASP.NET, Docker, PostgreSQL

Software Developer Intern - Alexa and Amazon devices

Amazon

Global technology company with businesses including e-commerce and cloud computing.

Built a Docker-based pipeline for synthesizing newly trained ML models, doubling generation speed.

• Developed pipeline for end to end synthesis of newly trained ML models which speeded up generation by 2x by packing all services into one using Docker • Tech Stack: C++, Python, Docker, AWS

Bachelor of Science, Computer Science

Nazarbayev University