Senior ML Research Engineer, NZ

Partly; Christchurch; On-site; NZD 100,000 - 130,000; Full time

Benefits offered by this job

Catered lunches

Annual wellness allowance

Paid parental leave

Commute allowance

Architectural-designed offices

Flexible work schedule

Team celebrations

Job summary

Partly is looking for a Senior ML Research Engineer in Christchurch to develop and deliver machine learning and algorithmic solutions for real problems in the vehicle and parts domain. You will collaborate with engineering teams, drive adoption and oversee the entire production rollout.

The ideal candidate will have strong experience in ML production, algorithmic thinking, and the ability to communicate effectively across teams. The role promises a vibrant office environment with ample wellness benefits and flexible working arrangements.

Qualifications

• Proven track record shipping ML into production.
• Strong algorithmic thinking with classical algorithms and data structures.
• Excellent applied modeling fundamentals and strong baselines.
• Evaluation-first mindset and clear articulation of failure modes.
• Engineering-minded execution focusing on maintainable code and reliability.
• Effective communicator and collaborator with stakeholders.
• Experience in messy, weakly-supervised domains.

Responsibilities

• Ship applied ML solutions end-to-end and own problem area.
• Design evaluation framework with gold datasets reflecting performance.
• Blend ML and algorithms pragmatically using the right tools.
• Build for production constraints considering latency and scale.
• Drive adoption across teams for impactful outcomes.
• Raise technical standards with reproducible experiments and documentation.

Skills

Machine Learning into Production

Algorithmic Thinking

Applied Modeling Fundamentals

Evaluation Framework Development

Engineering-minded Execution

Communication and Collaboration

Messy Weakly-supervised Domains

Search/Ranking/Retrieval

Job description

The Senior ML Research Engineer will build and ship machine‑learning and algorithmic solutions to real problems in the vehicle and parts domain. The role reports to Abram Spamers and works closely with engineering and product partners to take ambiguous inputs (noisy data, edge cases, shifting constraints) and turn them into measurable, production‑grade outcomes.

A core part of this role is helping to build a foundational model for the vehicle and parts problem space that can be adapted across multiple downstream tasks and product surfaces, and that improves over time as we expand data and evaluation. This role is for someone who wants to be judged by what ships: strong baselines, strong evaluation, reliable systems, and improvements that compound over time.

What will you do

• Ship applied ML solutions end-to-end. Own a problem area from framing through to production rollout, monitoring, and iteration.
• Design evaluation that makes progress undeniable. Build gold datasets and metrics that reflect real‑world performance.
• Blend ML and algorithms pragmatically. Use the right tools: modeling, ranking, classification, retrieval, graph/heuristic methods, LLMs, and domain‑specific algorithms where they outperform learning.
• Build for production constraints. Consider latency, scale, failure modes, observability, and safe rollout plans as part of the core deliverable.
• Work across teams to drive adoption. Partner with product and engineering so the solution actually changes outcomes, not just metrics.
• Raise technical standards by example: reproducible experiments, crisp documentation, thoughtful reviews, and clear trade‑offs that keep velocity high without breaking reliability.

Your skills

• Proven track record shipping ML into production. You’ve delivered systems used by others, and you understand monitoring, regressions, and operational realities.
• Strong algorithmic thinking. You’re comfortable with classical algorithms and data structures, and know when they beat ML.
• Excellent applied modeling fundamentals. You can build strong baselines, choose sensible methods, and evaluate correctly.
• Evaluation‑first mindset. You instinctively build the evaluation framework before you over‑invest in complexity, and you can articulate failure modes.
• Engineering‑minded execution. You write maintainable code, work effectively with services/pipelines, and care about performance and reliability.
• Clear communicator and collaborator. You can align stakeholders, document trade‑offs, and keep delivery moving in a low‑bureaucracy environment.
• Experience in messy, weakly‑supervised domains. You’ve worked where ground truth is imperfect and success requires clever measurement and iteration.
• (Bonus) Experience with search/ranking/retrieval or graph‑based approaches. You’ve built systems that combine multiple signals into reliable outputs.

Benefits

• Healthy, catered lunches—frequently fresh, healthy lunches every workday in our Auckland, Christchurch, London, and San Francisco offices.
• Healthy body, healthy mind—every team member gets a $1,500 annual wellness allowance on a Partly‑branded card to cover gym memberships, therapy, GP visits, and more.
• Family comes first—primary caregivers receive 3 months of fully paid parental leave, plus a flexible return‑to‑work (four days on full pay for your first three months back).
• Getting here is on us—paid 24/7 car park or commute allowance if you commute to a Partly office or co‑working space.
• Workspaces that inspire—architecture‑designed offices with coffee, social spaces, and nearby cafes.
• Office‑first with flexibility—default to in‑office daily in cities where we have a presence, while maintaining a high‑trust environment and flexible schedule.
• We celebrate together—weekly happy hours, monthly lunches, quarterly season openers, and an annual global offsite to keep team connection.