Applied Scientist / AI Research Engineer - Zinkworks

Zinkworks partners with leading Telecommunications and Financial Services organizations to modernize legacy systems, migrate mission-critical platforms to the cloud, and engineer AI-driven automation. From OSS transformation to rApp development and network intelligence, our teams simplify complexity and turn it into competitive advantage. Based in Ireland and operating across the EU, UK, and US, Zinkworks combines deep domain expertise with delivery excellence to help clients modernize faster and operate smarter.

Zinkworks is building a small product discovery team developing AI and machine learning capabilities for next-generation telecom networks, taking new product ideas from concept through to working prototypes.

We are looking for a senior applied scientist to lead the design of the machine learning methods behind these products. Working on complex problems in network data, you will determine the most appropriate modelling approaches, weigh the trade-offs, and build and validate the prototypes that prove them out. It is a hands-on role that combines research judgement with strong engineering practice, suited to someone equally comfortable selecting a method and implementing it.

Own method and model selection for the team's hardest ML problems, and defend the trade-offs with reasoning and evidence.

Turn ambiguous, real-world problems into sound ML approaches, and build the prototypes that test them.

Validate and challenge modelling decisions across the team; keep the bar high so "it should work" is always backed by evidence.

Adapt state-of-the-art techniques from the literature to the realities of messy operational data.

Work closely with domain experts to frame problems correctly, and with engineering colleagues to make prototypes production-ready.

Genuine depth in ML modelling: you can design, train, and critically evaluate models, not just call an API. Strong across several of: graph learning, spatio-temporal / time-series modelling, reinforcement learning, and causal inference.

A track record of taking a novel or ill-defined problem all the way to a working, validated model.

Strong Python and hands-on engineering; comfortable prototyping on cloud ML platforms (ideally GCP / Vertex AI).

The ability to justify method choices to technical and non-technical people, and to change your mind when the evidence says so.

Comfort in a fast-moving discovery environment with minimal support scaffolding, seeing problems through from question to working solution.

Telecoms, network-operations, or other complex real-time systems exposure.

Experience applying ML to large-scale operational or sensor/time-series data.

PhD or equivalent research depth in ML, paired with a record of shipping rather than pure academia.

Familiarity with LLM/agentic patterns as one tool among many, not the whole toolbox.