Photo by the author
# Entry
People often ask about my tech stack, specifically what I employ to build web applications, train machine learning models, and manage data science workflows. In miniature, I rely on a balanced mix of AI-based and other tools that enable me to work efficiently without sacrificing quality. These tools support everything from project planning and management to development, testing and implementation.
The best part? They are effortless to adopt. Most of them come with quick start guides, sensible default settings, and seamless integration with existing workflows, allowing you to incorporate them into your technology stack with minimal effort.
In this article, I will share seven imperative tools that can improve your workflow to a professional level. These tools will facilitate you become a better team member, a smarter developer, and a more effective developer from initial idea through production.
# 1. Git and GitHub: straightforward version control
Git is imperative for almost all developers and technical specialists. It helps you track code changes, debug and visualize project progress. You can even employ it to version models, datasets, and experiments. GitHub is the most popular platform that allows you to host projects and provides many tools and management features to facilitate you turn your ideas into production-ready designs in one place.
Why it’s great:
- Branching and Merging: Safely browse ideas on branches, then merge them when ready
- History and recovery: Exploit
git log,git diff,git stashAND reblog undo and restore - Pull requests and reviews: Discuss changes, perform checks, and maintain a pristine master branch
- GitHub activities: Automate tests, builds, and deployments with straightforward YAML
- Issues and projects: Track tasks, bugs, and roadmaps alongside your code
- Versions and packages: Mark versions, publish artifacts and manage change logs
- Security and Compliance: Dependency, code scanning, branch security and required reviews
I employ Git almost every day. Even when I’m involved in vibration coding, it’s a key part of my job. When I accidentally push unwanted changes or edit a previous commit, I employ Git to fix it. Trust me, I often submit a lot of unnecessary code and then realize I could have made simpler changes.
# 2. Cursor: AI-powered code editor
Cursor is a up-to-date editor built around artificial intelligence. It’s similar to VS Code, but adds a layer of intelligence that helps you write, repair, and refactor code faster. I believe this is a key tool for all coding problems. It now comes with multi-agent support, which means you can ask it to run multiple agents at the same time for collaborative troubleshooting. I employ it every day for coding, editing, autocomplete, and testing and pushing modern changes to projects.
Why it’s great:
- Built-in AI changes: Request changes directly to the file; get precise corrections in differential style
- Repo level context: Consider multiple files, symbols, and design architecture
- Multiple agent support: Break down problems and let coordinated agents handle subtasks
- Chat + Terminal Awareness: Reference logs, test results, and target fix commands
- Refactorers that stick: Secure name changes, interface changes, test generation and migration assistance
- Deep Git integration: Stage cuts, create approval messages and open PRs without leaving the editor
- VS Code ecosystem: Keep themes, keyboard shortcuts and most extensions
Many AI CLI tools provide integration with Cursor, allowing me to employ tools like Droid, ask them to build something for me, and see the changes in the Cursor IDE. It gives me control and helps me build things faster.
# 3. Claude Code: understands your entire project
Claude Koda is intended for developers working with huge code bases. It can read the entire repository and analyze multiple files at once. I really love Claude Code and I don’t even pay for the API or Claude plan. I employ it with the GLM encoding plan, which costs $3 a month and works better for me than any Claude Sonnet models.
Why it’s great:
- Repo-wide reasoning: Understands symbols, file dependencies, and architectural decisions
- Changes throughout the project: Proposes targeted differences/patches instead of throwing out walls of code
- Powerful scaffolding: Creates services, CLIs and templates with reasonable structure and documentation
- Testing and debugging: Generates unit/integration tests, tracks failures, and suggests fixes
- Tool Usage: Executes commands, reads/writes files, runs linters and checks logs on connected servers
- Documents and reviews: Summarizes modules, creates draft README files, and conducts thoughtful code reviews
Claude’s code is great for solving problems or creating modern applications. I used it to create a payment platform from scratch and it impresses with its capabilities. To get the most out of Claude Code, I highly recommend using the MCP Server, Claude Skills, and Claude Planning Discounts. Ask him to plan first and then execute.
# 4. Postman: Test your APIs with ease
Postman is a core set of API development tools. It makes it effortless to hit endpoints, check and visualize responses, and quickly debug. Even if you’re building a machine learning application, you still need to validate your inference and administrative endpoints. Postman provides a clear, visual picture of how your API is performing.
Why it’s great:
- Collections and environments: Organize requests, switch configurations (dev/stage/prod) using variables
- Built-in tests: Create quick JavaScript confirmations for status codes, payloads, and delays
- Monitors and automation: Plan your runs and get notified when something breaks
- Trial servers: Prototype endpoints before the backend is ready
- Cooperation: Share collections and documentation with your team with one click
There are many alternatives, and you can even create scripts for your own testers, but Postman stands out for its ease of employ, opulent feature set, and powerful collaboration tools.
# 5. Excalidraw: visualize your ideas
When you’re at a loss for words, sketch it out. Excalidraw makes it effortless to map system designs, workflows, and architectures, which is perfect for planning projects, blogs, presentations, or just thinking through a messy problem as it grows.
Why it’s great:
- Quick, hand-drawn effect: Communicate concepts without focusing on pixel-perfect details
- Shapes, connectors and labels: Perfect for flowcharts, ERDs, sequence diagrams and application maps
- Component libraries: Reuse UI templates, cloud icons and your own saved blocks
- Real-time collaboration: Brainstorm together, leave comments and repeat live
- Basic export and embedding: Drop diagrams on decks, documents or wikis (PNG/SVG/links)
# 6. Linear: Run your projects as planned
Linear provides speed and transparency in tracking issues. It’s swift, minimalistic and designed for engineering and product teams, perfect for planning content or delivering software without clutter. I employ Linear primarily for my work and I love it. You can assign tasks, share initial plans, and move items to different statuses. As you progress, you can view a history of changes and conversations, giving you a structured approach to content creation and project development.
Why it’s great:
- Lightning-fast UX and shortcuts: View triage, updates and searches.
- Issues, projects and cycles: Work structure from backlog → sprint → made with clear status flow.
- Custom workflows and labels: Customize statuses, priorities, SLAs and automations for your team.
- Deep integrations: Sync with GitHub/Bitbucket, connect PRs, receive Slack updates, attach projects and combine Notion documents.
- Real-time collaboration: Comments, mentions and activity schedules keep context in one place.
- Action plans and observations: Track your progress, speed and range changes at a glance.
# 7. Docker Desktop: Run anywhere, anytime
Docker makes your environment consistent. Package your application and all its dependencies so that it works the same on every computer, with no “works on my laptop” surprises. I employ Docker desktop for almost any project: local testing, rapid deployments and secure sandboxes for MLOps, data analytics, web development and testing modern AI models without touching my actual files.
Why it’s great:
- Repeatable environments: Send code and dependencies together as images for predictable flows
- Insulation and safety: Contains sandbox processes and file access so that experiments don’t leak onto your system
- Create for multi-service applications: Run APIs, databases, caches, and queues from a single docker
- Quick iteration: Layered builds, BuildKit, and caching speed up development loops
- GPU and machine learning support: Run CUDA/ROCm-enabled containers for local training/inference
- Multi-bow support and portability: Build for x86/ARM and deploy the same image to any cloud or on-premises
- Development containers: Standardize toolkits for your team in VS Code or JetBrains with a single setup
# Final thoughts
If you’re starting out as a developer or moving into a developer role, becoming proficient in these tools will facilitate you become faster and more proficient. You’ll be able to deliver features faster, collaborate better, and advance your career with confidence.
All the tools I mentioned are part of my daily toolkit: Git, Docker, Claude Code, Cursor, Excalidraw, and Linear. I employ them to create content and also to build machine learning and artificial intelligence applications.
I hope this article has given you a clear starting point and helped you choose the right tools for your coding journey.
Abid Ali Awan (@1abidaliawan) is a certified data science professional who loves building machine learning models. Currently, he focuses on creating content and writing technical blogs about machine learning and data science technologies. Abid holds a Master’s degree in Technology Management and a Bachelor’s degree in Telecommunications Engineering. His vision is to build an AI product using a graph neural network for students struggling with mental illness.