Technology
Proven tools, chosen for the job.
The languages, frameworks and platforms we use most, and the principle behind them: boring where it counts, modern where it pays.
What technology does Qodexify use?
Our defaults are React and Next.js on the front end, Node.js and Python on the back end, PostgreSQL for data and the major cloud platforms for hosting, plus language models, retrieval and MCP for AI work and n8n, Make or Zapier for automation. We pick proven, well-supported tools, add newer ones only where they clearly help and never lock you into a platform.
Our stack
What we build with
Front end
React, Next.js, TypeScript, Tailwind and modern CSS, with a focus on speed, accessibility and SEO.
React · Next.js · TypeScript
Back end
Node.js and Python services with clean APIs, queues and background jobs.
Node.js · Python · REST and GraphQL
Data
PostgreSQL as the default, with Redis for caching and search or vector indexes where needed.
PostgreSQL · Redis · Vector search
AI and agents
Hosted and open language models, RAG, tool use, MCP servers, evaluation and monitoring.
LLM APIs · RAG · MCP
Automation
n8n, Make and Zapier for fast wins, and custom workflows when you outgrow them.
n8n · Make · Zapier
Cloud and delivery
AWS, Google Cloud or Azure with containers, infrastructure as code, CI/CD and monitoring.
Containers · CI/CD · Observability
Principles
How we choose technology
Maturity over novelty
Well-supported tools mean fewer surprises, easier hiring and cheaper maintenance.
Portable by design
Open standards and clean architecture, so you can move providers or bring work in-house.
Security as a default
Least privilege, patching, secrets management and dependency checks, as covered in our supply chain guide.
Measured performance
Speed and accessibility are tested, not assumed. See our Core Web Vitals checklist.
AI with guardrails
Evaluation, monitoring and approvals on every AI feature. See our guardrails guide.
Fit for your team
We work in your stack where it makes sense and document everything we add.
FAQ
Frequently asked questions
Do you only use these tools?
No. These are our defaults because they are mature and well supported. We choose the right tool for each job, and we are happy to work in your existing stack.
Why do you recommend boring technology?
Proven tools are faster, safer and easier to hire for and maintain. We use newer tools where they give a clear advantage.
Do you lock us into a platform?
No. We favour open standards and portable designs, and hand over code, documentation and infrastructure access.
Which AI models do you use?
We choose per task from leading hosted models and smaller or private models, based on quality, cost, speed and data requirements.
Can you work with our legacy systems?
Yes. We integrate with and gradually modernise older systems. See our legacy modernization service.