Client work · Sofia, Bulgaria
Intelligent Aggregator Platform
Ask the catalogue a question, get a real answer.
A catalogue of B2B software where buyers ask a question in plain English and get an answer back from the live listings, instead of clicking through category pages.
- Role
- Full-stack engineer
- Timeline
- Jul 2023 - Nov 2024
- Scope
- Database design, rendering, search, AI agent
- Client
- Not named publicly
01
Inside the build
Rendering and routing
Three rendering modes share one site. Listing and vendor pages are pre-rendered at build time so they load instantly and crawl cleanly. Incremental regeneration refreshes those pages in the background when the underlying rows change, so nobody waits for a rebuild. Pages that depend on a live query are server-rendered on request.
Routes are generated from the database, and so are the sitemaps and URL structure. Caching sits in layers in front of PostgreSQL, so repeat traffic on popular pages keeps database load low.
The AI layer
Two pieces sit on top of the same schema.
Semantic search runs on pgvector. Embeddings live next to the product columns, so one query can rank by meaning and filter by category at the same time.
The agent handles the questions a filter cannot answer. It reads the question, chooses the queries, calls PostgreSQL through tool calling, and returns a report built from the current data.
Interface and data
On the page
- Empty states tell users why there is nothing here and what to do next
- CTA placement was A/B tested rather than argued about
- WCAG compliance, so keyboard and screen reader users get the same catalogue
- Vendor profiles detailed enough to make an informed decision
- Ads placed off the reading path, so they do not fight the product
Behind it
- Zod schemas validate data at every boundary
- Supabase Storage serves images, compressed on the way in
- Drizzle keeps queries type-safe against the same schema the app types come from
- Feedback channels built into the product, so problems get reported instead of guessed at
Stack
Next.js and TypeScript on the front, Tailwind and shadcn/ui for the interface. Drizzle over Supabase PostgreSQL with pgvector for storage and search. OpenAI for query understanding and embeddings. Deployed on Vercel.
02
Related work
Outrizz AI: Event Engagement Platform
30K+leads generated
Outrizz AI: Event Engagement Platform
An event marketing platform for B2B sales teams. It finds the right people before a trade show and talks them into coming, then captures everyone who stops at the booth and follows them up while the event is still running.
Next.jsReact NativeAItRPCTurbo MonorepoDrizzleSupabase
View Project
Recruit AI: Intelligent Recruitment Matching
Recruit AI: Intelligent Recruitment Matching
An open-source hiring marketplace where the ranking is the product: jobs surface the right candidates, candidates surface the right jobs, and both sides are scored by the same rules. Built as a reference for search that gets measured instead of guessed at.
Next.jsEffect.tsQdrantVector SearchRetrieve & RerankVercel AI SDKDrizzlePostgreSQLTurborepo
View Project
PickArt: Art-Hospitality Marketplace Platform
15Kmonthly active users
PickArt: Art-Hospitality Marketplace Platform
A marketplace that puts artists' work on hotel walls and lets guests buy the piece they are looking at: hosts get free curation, artists get a sales channel, guests take the art home.
Next.jsTypeScriptSupabaseStripeReact PDFNext-intlDrizzleTanStack Query
View Project
Building a catalogue or directory product?
Directory products live or die on discovery: whether a buyer can find the right row, and whether Google can find the page. Tell me what you are cataloguing and I'll show you the fastest path to both.
Book a free call30 minutes, free. You leave with a plan either way.