Junior Data & Machine Learning Specialist (Contract)

Stichting Mission on Wheels; Calgary; Hybrid; CAD 55,000 - 60,000; Full time

Benefits offered by this job

Health Spending Account

Calgary office

Employee discounts on gear

Staff gym access

Team that likes what they do

Job summary

Stichting Mission on Wheels is seeking a builder to ship a training app and turn raw sensor data into trusted datasets. This hybrid Calgary-based role collaborates with engineers, coaches and the product team.

The contract is six months, with potential to go full-time. CAD 55,000–60,000 base plus Health Spending Account, staff gym access and gear discounts. Core hours Monday–Friday 8:00–16:30, hybrid on-site three days a week.

Qualifications

• Strong Python programming and data wrangling skills.
• Experience with SQL and version control (Git).
• Experience with iOS and Android development.
• Ability to explain analyses and reproducibility of results.
• Capable of building and maintaining data pipelines.

Responsibilities

• Build Python/SQL tools to ingest and validate sensor data and workout logs.
• Ensure datasets are traceable with IDs, equipment, timestamps and consent.
• Run collection checklists with the team to catch missing samples and mislabels.
• Create baselines and reproducible ML experiments with proper train/test splits.
• Document errors, limitations, and provide interfaces for the app team.

Skills

Python

Data wrangling

SQL

Git

iOS development

Android development

Pandas/NumPy

scikit-learn

PyTorch

Education

Relevant degree or diploma

Tools

Versioned data pipelines

Job description

About the role

We’re looking for a builder who wants to ship an app people actually train with.

We’re building a training app that makes the gear people already own more useful. Alongside it we’re running a wearable research program testing whether affordable sensors can give people better coaching. You’ll take raw sensor movement recordings and turn them into datasets we can trust and findings we can test. You’ll work with the founder, our engineers, our coaches and the full-stack developer.

You don’t need years of experience. You need to care about getting the data right and be honest about what a result does and doesn’t prove.

What’s in it for you

• This is a 6-month contract role ending in March 2027, with the potential to transition to full-time based on performance and business needs
• CAD $55,000–$60,000 base salary
• Health Spending Account
• A Calgary office you’ll actually want to come into
• Employee discounts on all our gear
• Access to our staff gym
• A team that actually likes what they do

The schedule & location

This is a hybrid role based at our Calgary location, where the core hours are Monday to Friday from 8:00am to 4:30pm

The non-negotiables

You build with Claude Code every day, and you check its work. You can explain every script and analysis you hand us, and you’ve verified the results hold up. Fast matters. Shipping slop doesn’t.

You lift. You know sets, reps, RPE, progression and rest periods, and you’re at home around barbells, racks, benches, dumbbells and cable machines. No coaching cert needed. You can’t label a squat rep well if you’ve never done one.

You’re in Calgary. You live here or you’ll move here before your start date. The job is hybrid but you are expected to come on-site around 3 times a week to work with our equipment, coaches and product team, so remote from another city or country won’t work.

What you’ll do

• Build and maintain Python/SQL tools that pull in and check sensor recordings, workout logs and reference measurements
• Keep every dataset traceable: participant IDs, exercise, gym and equipment setup, sensor placement, timestamps, consent, labels and version
• Run collection checklists with the testing team, and catch missing samples, timing errors and bad labels before they reach an experiment
• Build simple baselines and documented ML experiments, with reproducible configs and evaluation on properly separated users and sessions
• Write up errors and limitations, keep the negative results, and maintain records and interfaces the app team can use

What Success Looks Like

• Day 30: you’ve reproduced an existing analysis and built a checked dataset for one exercise and sensor setup
• Day 60: reference-data checks and baseline comparisons are running under an approved protocol
• Day 90: you own a reproducible dataset-to-evaluation workflow and a clear error report that feeds the next sensor decision
• Nothing becomes coaching in the app until it’s been validated, and you’re the reason we know

What We’re Looking For

• Practical Python, data wrangling, SQL and Git. You can turn a notebook into a repeatable script and write checks for your own assumptions
• A foundation in stats and ML: keeping training and validation data separate, spotting overfitting, comparing against a baseline
• At least one real project, from a job, research, a capstone or something substantial you built on your own. New grads and people with 1–3 years of experience welcome
• Patience with messy data, curiosity about where things break, and the ability to explain what a result does and doesn’t show
• Experience with iOS and Android development

Bonus Points

• Time-series analysis, signal processing, wearable sensors or biomechanics
• pandas/NumPy, scikit-learn or PyTorch
• Versioned data pipelines
• A relevant degree or diploma (equivalent ability counts just as much). No PhD or production experience required. Coaches and engineers support the reference sessions and help interpret results.

_ We care more about what you’ve built than where you learned it. _

The Interview Process

• 30 minute screening with our HR Coordinator
• 1 hour screening with our hiring team
• Paid, time-bound technical assessment
• 1 hour technical follow-up to discuss assessment
• Decision made and offer sent out