The Problem

Online shopping isn't shopping anymore.
It's a scavenger hunt.

Finding the best-value item online today is inefficient and fragmented. You jump between Shopee, Lazada, Amazon, and Carousell like a man possessed, opening tabs, guessing keywords, and second-guessing yourself.

The Solution
Shoppo circular icon

Meet Shoppo.

Shoppo is a Telegram-native intelligent shopping agent. Users can start with messy intent like text, a budget, or a product image inside one chat thread. Shoppo clarifies what they actually want, uses TinyFish to fan search across marketplaces, and returns a ranked shortlist with evidence instead of making users manually hunt across tabs.

Telegram logo
User Surface

Telegram Native

Everything stays in one familiar chat. Users can send rough requests, photos, clarifications, and follow-up constraints without jumping into another app or portal.

TinyFish icon
Execution Layer

TinyFish Powered

Once the brief is ready, TinyFish runs the search across marketplaces in parallel, preserving search coverage, raw evidence, and execution transparency.

Shoppo icon
Outcome

Decision Support

Shoppo does not just retrieve listings. It helps users decide by surfacing the strongest candidates, trade-offs, and best-value options in a clearer final shortlist.

Text / Photo / Budget -> Telegram Chat -> Shoppo Agent -> TinyFish Search Fan-out -> Ranked Shortlist

Watch Shoppo in Action

This is the end-to-end moment: messy user intent goes into Telegram, TinyFish fans search out across marketplaces, and Shoppo returns a clearer shortlist instead of a tab jungle.

How Shoppo actually resolves the hunt.

01
🔍
Cross-Platform

Shoppo automatically searches across all major platforms simultaneously. No more tab-hopping or repeating the exact same search.

02
🧠
Query Optimization

Most users start with a messy, vague idea. Our agent actively guides you to clarify your intent and translates it into the optimal search query.

03
📊
Logical Ranking

We don't just dump raw links. Our decision engine evaluates the results and ranks the absolute best candidates based on price, value, and relevance.

04
🤖
Telegram Native

Absolutely zero friction. This entire pipeline operates completely inside a single Telegram bot. No new apps to download, no messy web portals.

Conversation is the right front door.

The first core idea is that shopping intent begins messy. Users do not start with the exact keyword a marketplace expects. They start with fragments. This slide diagrams why Telegram plus clarification is the right intake surface before the system ever spends search effort.

Text Request Reference Photo Budget Range Follow-up Clarification
Telegram logo
User Surface

Telegram Thread

All messy intent starts in one familiar place instead of forcing the shopper into a rigid search form or another portal.

->
Shoppo icon
Agent Layer

Shoppo Clarifies

The agent resolves ambiguity, asks for the missing detail, and turns a vague request into something precise enough to search well.

->
Shoppo icon
Search Input

Structured Brief

The output of conversation is not more chat. It is a clarified sourcing brief that can now drive deterministic and grounded search.

Principle

Natural Input

The user should be allowed to describe the product the way people actually think, not the way search engines expect.

Principle

Context Preservation

Because the workflow stays in one Telegram thread, every clarification, image, and constraint remains part of the same search context.

Principle

Zero Friction

The shopper does not need a new app, a new portal, or a new behavior. They just chat the way they already do.

The product is confidence, not more links.

The second core idea is that retrieval is not the end product. Confidence is. This diagram shows the pipeline that creates that confidence: marketplace fan-out through TinyFish, preserved evidence, and a final decision layer that turns raw search output into a trustworthy shortlist.

TinyFish icon
Execution Layer

TinyFish Orchestrator

TinyFish is what turns one user request into real multi-marketplace execution rather than a single brittle search attempt.

->
Marketplace cluster visual
Coverage

Marketplace Fan-out

Search is broadened across marketplaces so the shopper is not left repeating the same manual comparison work tab by tab.

Carousell Lazada Amazon SG Alibaba
->
Shoppo icon
Decision Layer

Evidence + Ranking

Shoppo preserves evidence, compares trade-offs, and ranks the strongest candidates so the answer is explainable instead of opaque.

->
Shoppo icon
User Outcome

Confident Output

The shopper receives something actionable and trustworthy, not just another list of raw links.

Telegram Shortlist Live Evidence Report
Principle

Coverage

The system should broaden the search space for the user, not leave them repeating manual searches across disconnected platforms.

Principle

Transparency

Confidence comes from preserved evidence, visible trade-offs, and a workflow that does not feel like an opaque black box.

Principle

Decision Support

The final output should help the user choose, not just click. That is why Shoppo ranks, explains, and narrows the shortlist.