The infrastructure our products ship on.
An opinionated, boring, production-grade stack behind every product we build. We pick tools that survive without supervision and we don't add a layer unless there's a reason to.
Tools by layer
| Layer | Tools |
|---|---|
| Mobile | Android (Kotlin) · Capacitor |
| LLMs | Google Gemini · Anthropic Claude |
| Embeddings / RAG | Supabase pgvector · Gemini embeddings |
| Storage | Supabase — Postgres + pgvector |
| Edge / Deploy | Cloudflare Workers · R2 · D1 |
| Web | Vite · React |
| Payments | Razorpay · Stripe |
| Orchestration | n8n (self-hosted) |
Why this stack
Android with Kotlin is where ToTa-Ai lives, on-device and close to the notification stream. Capacitor lets Swaad CMS ship a mobile-first product without a separate native codebase. Cloudflare Workers, R2, and D1 give freeshare.biz a global edge with zero servers to run. Supabase + pgvector replaces three tools at once — auth, Postgres, and a vector store — and is the backbone of the KRS RAG corpus.
Gemini is the default for embeddings and high-volume inference because of price and a usable API. Claude is the default when reasoning quality matters more than throughput. Razorpay and Stripe handle payments; n8n, self-hosted, handles the glue we'd otherwise rebuild.
Philosophy
We build products, not slideware. Every product we ship comes from a problem we've actually lived. If it needs a deck to explain why it matters, we didn't build the right thing.
What we won't add
— Vector DBs that aren't pgvector (we don't need a separate service)
— LangChain wrappers around things that are 20 lines of code
— Vendor agent platforms — we build the orchestration ourselves
— Frontend frameworks for marketing pages (this site is vanilla HTML)
Want this stack pointed at your vertical?
Product partnerships and async consulting where we've already solved the problem.
Get in touch →