BuyWhere
← Back to documentation

linkedin_posts

LinkedIn Posts — BuyWhere Founder-Voice Distribution

Post 1: Founder Announcement — "We Built BuyWhere for AI Agents"

Timing: Week 1 (April 14–18) UTM: ?utm_source=linkedin&utm_medium=social&utm_campaign=linkedin-founder-m1

Body:

Six months ago, I asked our AI assistant to find the cheapest 65-inch OLED TV in Singapore.

It hallucinated three prices. None were real.

That's when I understood the real problem with AI shopping assistants: the product data layer doesn't exist for AI agents. Price comparison sites are built for humans. Official merchant APIs require business registration. Scraping works until it doesn't.

So we built BuyWhere — a product catalog API for Singapore e-commerce, designed from day one for AI agent consumption.

What that means practically:

  • Structured JSON-LD responses that LLMs parse without instruction
  • MCP server so your Claude/GPT-4 agent can search products as a tool call
  • Real-time prices from Lazada, Shopee, Best Denki, Courts, Harvey Norman, Qoo10, and 45+ other merchants
  • 1M+ products indexed and updated on a rolling schedule

The future of commerce isn't a chatbot. It's an agent that can actually check live prices, compare options, and give you an affiliate link.

We're still early. But if you're building shopping agents, deal bots, or any commerce layer on top of LLMs — I'd love your feedback.

Free API access at buywhere.ai/api-keys

What's broken about AI shopping assistants right now? Genuinely curious what others are running into.

Post 2: Developer Angle — "Structured Data for AI Agents"

Timing: Week 2 (April 21–25) UTM: ?utm_source=linkedin&utm_medium=social&utm_campaign=linkedin-dev-m1

Body:

Most AI shopping tools are chatbot wrappers. You have a conversation, and somewhere in the background, someone is maintaining scrapers that break every time the e-commerce platform changes their HTML.

We took a different approach with BuyWhere.

Instead of building another chatbot, we built the data layer: a product catalog API that serves Singapore e-commerce in Schema.org JSON-LD format — the markup that LLMs were trained on.

What that means:

  • No custom parsing instructions needed
  • Agents get price, merchant, availability, and affiliate link in a format they already understand
  • MCP server for zero-config tool registration in Claude Desktop

Coverage: Lazada, Shopee, Best Denki, Courts, Harvey Norman, Qoo10, Gain City, iStudio, and 45+ more merchants. 1M+ products.

If you're building anything that involves product recommendations, price comparison, or shopping assistance with AI — the API might be useful.

Docs: buywhere.ai/api-keys

Post 3: Singapore Market — "Why E-Commerce Data is Broken in SEA"

Timing: Week 3 (April 28 – May 2) UTM: ?utm_source=linkedin&utm_medium=social&utm_campaign=linkedin-sg-m1

Body:

Singapore has world-class e-commerce infrastructure — but no good price comparison layer on top of it.

You want to buy a laptop. You open Shopee, Lazada, Best Denki, Courts, Harvey Norman, and Qoo10. Five tabs. Five different experiences. Five different definitions of "in stock."

For humans, it's annoying. For AI agents, it's a data engineering nightmare.

This is the problem we're solving at BuyWhere. One API, normalized product data, Schema.org structured responses — so your agent can actually comparison shop without you building a scraping pipeline for every Singapore retailer.

Currently covering 52 merchants and 1M+ products. Updated continuously.

If you run any kind of commerce or deal-finding product in Singapore, I'd love to compare notes on data challenges.

Note: These are drafts. Personalize with your own voice before posting. LinkedIn rewards authenticity over polish.