R0009683: (TrendLife) (Sr.) AI Engineer

ABOUT TRENDLIFE

TrendLife is our consumer-focused division, delivering the best security, privacy, and anti-scam solutions for everyday users. We build products that protect real people from real threats — phishing, identity fraud, malicious content, and more — and we do it by combining deep security expertise with fast, iterative product development.

[ABOUT THIS ROLE]

In a domain as fast-moving as consumer security and anti-scam, the ability to learn quickly, adapt, and validate ideas through real prototypes is more valuable than a perfect resume. This role is for someone who uses AI to move faster, thinks beyond the obvious solution, and is comfortable making real things under ambiguity.

The scope of this role spans both ends of the spectrum: building new products and services from the ground up, and rethinking or modifying the architecture of existing ones. Both happen here, and you will need to be comfortable with both.

WHAT WE ARE LOOKING FOR

-AI-augmented development
-You actively use AI tools to accelerate your own development work — not just for code completion, but to tackle harder problems faster, explore unfamiliar domains, and deliver more in less time. This applies across product development, service delivery, and internal tooling. We are not prescriptive about language — C, C++, Java, Python, Go, or anything else. If you can build a convincing PoC in an unfamiliar language because you know how to work effectively with AI, that is exactly the point.
e.g. Using an LLM to rapidly prototype a scam-detection classifier, then iterating with real user feedback — rather than waiting for a complete spec before writing any code.
-When you hit a problem, you dig into the underlying mechanism rather than just patching the surface. In the security domain especially, this matters: understanding why a threat evades detection, or why a model behaves unexpectedly, is often the difference between a real fix and a temporary workaround.
e.g. Investigating why an anti-phishing model produces false positives on certain URL patterns — tracing it back to training data distribution rather than just adjusting thresholds.
-Comfort with ambiguity and change
-The AI and threat landscape moves faster than any roadmap can anticipate. You do not need a fully defined spec to get started. You can take a vague problem, make some reasonable assumptions, build a quick prototype, and use the result to sharpen the question. New attack vectors, shifting user behaviors, and evolving AI capabilities are not disruptions to you — they are the environment you work in.
e.g. When a new type of voice-cloning scam emerges, jumping in to explore detection approaches before the full product requirement is written.
-POC-first work habit
-You are comfortable working in rapid cycles: rough prototype, real feedback, next iteration. This startup-style cadence feels natural to you, and you have a track record (even from side projects or academic work) of shipping things quickly to validate ideas.
e.g. Spinning up a lightweight demo of an AI-powered privacy scanner in a few days to pressure-test assumptions with the team — before committing to a full implementation plan.

[NICE TO HAVE]

-Follows new tech
-Keeps up with developments in AI and adjacent fields — not just passively, but with enough judgment to separate signal from noise. Can spot which emerging capability (e.g. multimodal models, on-device AI) might be relevant to consumer security before it becomes obvious.
-Business awareness
-Has some sense of how a product creates value and why certain technical choices matter commercially. Can connect an engineering decision to a real-world outcome