AI in our software development
Where AI sits in our work
Discovery
We read client briefs, existing codebases, and reference systems in hours instead of days. That speed gets passed back to you in faster proposals and earlier decisions.
Architecture
AI generates alternative designs and surfaces tradeoffs our architects might not have considered. Better decisions, made sooner.
Coding
Boilerplate, scaffolding, refactors, and test stubs move at AI speed. The team focuses on the parts of the codebase that need thought.
Testing
AI generates edge-case coverage that often gets skipped under deadline pressure. Your software ships tested against scenarios the spec didn't list
Documentation
Inline comments, READMEs, and API references generated alongside the code and checked at handover. Documentation that exists.
Project management
Our AI project assistant, Owtoo, reads from our PM tools, repos, and meeting notes, then answers status questions and surfaces blockers. PMs spend their time on the blockers themselves.
The models we use
We don’t standardize on one provider. Different models suit different work, and the field moves too fast to lock in.
OpenAI. General drafting, structured outputs, tool calling. The default when the job isn’t obviously something else.
Codex. Code-heavy tasks. Repo-aware refactors, migration scripts, PR review assistance.
Claude (Opus and Sonnet). Long documents, careful reasoning, work that needs to stay inside specified constraints across a long run.
Manus. Open-ended autonomous tasks. Research, multi-source comparisons, document assembly.
Our own AI products
We build AI software for clients. We also use it ourselves.
Owtoo
Internal project assistant. Reads from our project management tools, documents, and code repositories. Answers status questions, drafts updates, surfaces blockers.
WriteText.ai
Our ecommerce content product. It runs the full content loop inside your store: keyword analysis, content generation across descriptions, meta fields, alt text, FAQs, PAA blocks, and category pages, multilingual delivery in dozens of languages, then automatic rewrites when rankings move. Built native into WooCommerce, Magento, and Shopify, with an API for everything else.
What this means for your project
Discovery moves faster
We digest your existing systems and documentation faster, which means less time explaining context and more time on important decisions.
Documentation arrives with the code
Generated during development, reviewed before handover. Your project doesn’t ship with a documentation debt.
AI-ready architecture, when it fits
If your product could plausibly use AI features later (search, content, automation, support), we design so adding them doesn’t mean rebuilding.
Ready to explore AI for your business?
Whether you want to automate customer support, optimize internal workflows, or build AI-powered features into your products—let’s discuss what’s actually possible for your business.
Got a problem? Let’s brainstorm if AI can solve it.
Commonly asked questions
Does AI in your process make my project cheaper?
Sometimes. More often it shows up as more scope in the same budget.
Will my code be generated by AI?
Parts of it. AI handles boilerplate, scaffolding, refactors, and test stubs at speed.
Is my code or data sent to external AI providers?
Depends on the project. For most work we use commercial AI tools with data-retention-off settings and no training on customer data. For projects with strict residency or confidentiality requirements, we set the deployment to match.
How do you decide when to use AI?
The test is whether AI makes the work better, not just faster. Faster but worse isn’t progress.
Can you build AI features into the product you're making for me?
Yes. That’s covered on our AI software services page.
Book a consultation to explore how AI and 1902 Software can transform your next ecommerce project — faster, smarter, and more cost-efficient.