Senior practitioners who ship production systems.
Founded by engineers who have built enterprise-scale AI and data platforms, we help organizations turn complex data into practical solutions—then hand them back to your team.
Years combined experience
PhDs in physics
Daily listings processed
Messages classified
Meet the engineers.
Data engineers and ML practitioners who have built systems at scale for enterprise organizations.
Solomon Owerre, Ph.D.
Founding ML Engineer
Former Director of Data Science at Loblaw and Sr. ML Engineer at Wiser Solutions. Delivered fraud-detection work that generated $8M in annual savings and built multilingual product-classification systems processing 17.5M+ listings daily.
Joachim Nsofini, Ph.D.
Founding Data Scientist
Senior Data Scientist at FCC / FAC Canada. Built a relationship-manager text classifier with 95% accuracy across 1M+ messages, and analyzed 1M+ relief-program records to support funding-integrity decisions.
A transparent engagement model.
Each step has a written deliverable and a stop point. If the next phase isn't the right move, we say so.
Discover
Two weeks. We sit with your data, team, and KPIs. Output: a 1-page scope, a baseline, and a go/no-go recommendation.
Design
Architecture, data contracts, evaluation plan, and rollout strategy — reviewed with your engineers before any code lands.
Build
Fixed-fee delivery in your repo and your CI. Weekly demos. The work merges as it's reviewed — no big-bang handoff.
Optimize
Post-launch tuning, monitoring, and a runbook. Then we step out — or stay on retainer if that's what the system needs.
What we won't compromise on.
Smallest change first
We propose the narrowest scope that moves the metric. Bigger engagements happen after the first one earns the next.
Reviewable artifacts
Pull requests, design docs, eval reports. Nothing important lives in a private notebook.
No lock-in
We use the stack you already pay for. When we leave, your team can extend the system without us.
Tools we work with daily.
Ready to scope the first phase?
A 30-minute call is usually enough to know whether we're a fit. No sales pitch — just a technical conversation.