Skip to content
@TankDev-tech

TankDev.tech

Software Engineering · Applied AI · Automation

TankDev

Software Engineering · Applied AI · Process Automation

TankDev is a software engineering organization focused on building custom software, applied artificial intelligence systems, and automation for real operational workflows.

We work across the full engineering lifecycle — from problem analysis and system architecture to implementation, integration, deployment, and operational tooling.

Engineering Focus

Custom Software
Purpose-built software for business and operational requirements that are not adequately served by off-the-shelf tools.

Applied AI
Machine learning and AI systems designed around measurable technical objectives, explainability, and real implementation constraints.

Process Automation
Software that transforms repetitive, fragmented, or manual workflows into controlled and traceable systems.

Web Applications
Modern web platforms combining user-facing interfaces, backend services, databases, administration tools, and external integrations.

System Integration
Connecting software, APIs, databases, desktop tools, and existing operational systems into unified workflows.

Selected Engineering Work

DeepCancer

Explainable AI for Breast Histopathology Classification

A research project combining Swin Transformer and ConvNeXt architectures for breast histopathology image classification, with Grad-CAM explainability.

The published study evaluates binary and eight-class classification on the BreaKHis dataset. The binary experiment reported 98.86% accuracy, 99.07% precision, 99.25% recall, and 99.16% F1-score.

Research & education only — not for clinical diagnosis.

GitHub Repository · Research Platform · System Overview · Case Study · Published Research


Web-to-Print Production Automation

From browser-based personalization to production-ready print files

A custom production system connecting customer-facing design workflows, order management, production operations, and Adobe Illustrator-based output generation.

The system brings multiple stages of a print-production workflow into a coordinated software architecture using technologies including Next.js, FastAPI, and PostgreSQL.

GitHub Repository · System Overview · Case Study


Autonomous AI Video Production System

Research-to-publishing AI production infrastructure

An autonomous production system coordinating research, source processing, script generation, voice production, visual generation, media assembly, production controls, metadata, scheduling, and publishing.

The architecture includes 8 coordinated production stages, 4 API budget control levels, 2 independent production flows, and a centralized operational control layer.

GitHub Repository · System Overview · Case Study

Engineering Approach

TankDev systems are built around a few consistent principles:

  • Solve the operational problem before choosing the technology
  • Prefer maintainable system architecture over unnecessary complexity
  • Automate repetitive work while preserving control and observability
  • Treat AI as an engineering component rather than a standalone feature
  • Design around real deployment and infrastructure constraints
  • Document system boundaries and avoid unsupported performance claims
  • Keep public technical documentation separate from proprietary implementation

Technology

Our work commonly involves:

TypeScript · Python · Next.js · FastAPI · PostgreSQL · PyTorch · Tailwind CSS · Linux · REST APIs · Playwright

Technology choices vary according to the requirements of each system.

Public Engineering Documentation

The repositories published under this organization are primarily technical showcases and engineering documentation.

Depending on the project, public repositories may document:

  • System architecture
  • Engineering decisions
  • Production workflows
  • Experimental methodology
  • Research results
  • Interface behavior
  • Integration boundaries

Production source code, customer data, credentials, private infrastructure, model weights, proprietary business logic, and security-sensitive configuration are not published unless explicitly intended for open-source distribution.

Research

Histopathological Breast Cancer Classification Using a Hybrid Swin Transformer and ConvNeXt Architecture

Murat Onur Kaderoğlu · Emre Şatır
Journal of Artificial Intelligence with Applications, 2025

DOI: 10.5281/zenodo.18138734

TankDev

Website: tankdev.tech
Case Studies: tankdev.tech/tr/case-studies
GitHub: github.com/TankDev-tech
LinkedIn: TankDev


Custom Software · Applied AI · Process Automation · System Integration

TankDev.tech

Pinned Loading

  1. web-to-print-automation web-to-print-automation Public

    Technical showcase of a custom web-to-print production automation system developed by TankDev.

  2. ai-video-production-system ai-video-production-system Public

    Technical showcase of an autonomous AI-powered video production and publishing system developed by TankDev.

  3. deepcancer deepcancer Public

    Explainable AI research prototype for breast histopathology classification using Swin Transformer, ConvNeXt, and Grad-CAM.

Repositories

Showing 4 of 4 repositories

Top languages

Loading…

Most used topics

Loading…