Mlytics Careers

We turn content into commerce. The AI Answer Engine for the intent economy.

We’re building the Intent Refinery — a platform that captures reader intent at the point of content consumption, matches it to brands, and converts it into measurable commercial outcomes. Not impressions. Not clicks. Declared purchase intent, priced on CPL.

About Mlytics

Mlytics is an AI-powered Answer Engine built for the intent economy.

We help media publishers turn reader intent into measurable commercial outcomes — replacing fading CPM ad revenue with high-quality CPL revenue. We help brands stop buying impressions and start buying confirmed intent.

Our 5-layer Intent Refinery — Content → Decisive Engine → AI Q&A → Full Conversation → Lead Pilot — is live with 15+ of Taiwan’s top media properties and 4M+ weekly active users.

We started as a multi-CDN company. That infrastructure — <50ms routing, multi-vendor failover — is now the substrate underneath. Same engineering rigor, much bigger ambition.

By the numbers

4M+ weekly active users across the platform
15+ live publisher integrations (Taiwan’s top media properties)
2,500x efficiency gain: $0.10 AI-generated content vs. $250 human-written

What we work on

Real distributed systems at meaningful scale. 4M+ WAU, <50ms p95, multi-vendor failover (Cloudflare / Akamai / Fastly), real CDN economics.
AI agents that actually do work. Intent classification, RAG over publisher-specific knowledge bases, multi-step agent runtime. Not chatbot demos.
Two-sided marketplace mechanics. Publisher supply meets brand demand through real-time intent matching. CPL pricing. Editorial guardrails. Real money flowing.
Product engineering, not feature engineering. Every engineer participates in scoping. Every slice has a paying customer gate. If it doesn’t ship to a real user, it didn’t ship.

Open Positions

Role	Location	Team
AI-Powered Full Stack Engineer	Singapore / Taipei	Product
AI Data Engineer	Singapore / Taipei	Data & Innovation

How we ship

In early April 2026, a team of four shipped a full AI video news pipeline in six hours. Nine days later, one engineer paired with Claude Opus shipped a two-agent brand-integration system in a day. Across our three newest production codebases, over a quarter of all commits list Claude as a co-author. This is what AI-first engineering looks like from the inside, not the press release.

How we ship → Becoming Product Builders with Business Thinking

How to Apply

Send us something that shows how you think — a side project, a technical write-up, a PR you’re proud of, or a message explaining why this caught your attention.

📧 [email protected]

We care about what you’ve built more than where you’ve worked.

Questions? → Candidate FAQ