01
Data platform & pipeline engineering
We design and implement batch and streaming pipelines that land curated datasets in your warehouse or lakehouse. Work includes source connectors, schema contracts, data quality checks, orchestration (Airflow, Dagster, or native cloud schedulers), and cost-aware partitioning. Ideal when analytics teams wait days for fresh tables or when legacy ETL scripts nobody dares to touch block new products. Engagements often start with a lineage audit and finish with tested jobs your engineers can extend.
Typical range: CAD $18,000 – $95,000 per project
02
Applied machine learning
We build supervised and unsupervised models for forecasting, classification, ranking, and anomaly detection tied to measurable business outcomes. Feature engineering draws from your existing stores—not synthetic demos. We document training data boundaries, evaluation metrics, and known failure modes. Models ship with inference endpoints or batch scoring jobs plus monitoring hooks. We do not promise perfect accuracy; we promise reproducible experiments and honest performance reporting your stakeholders can scrutinise.
Typical range: CAD $25,000 – $120,000 per use case
03
LLM integration & RAG systems
When generative AI fits, we implement retrieval-augmented generation over your documents, tickets, or product catalogues with access controls that mirror existing permissions. Work covers chunking strategy, embedding pipelines, vector stores, prompt templates, guardrails, and human-in-the-loop review for high-risk answers. We benchmark latency and hallucination rates against baselines—not marketing claims—and advise when a simpler search index would suffice.
Typical range: CAD $30,000 – $140,000 per deployment
04
MLOps & model operations
We operationalise the path from notebook to production: CI/CD for training code, model registries, canary releases, drift detection, and on-call runbooks. Integrations with MLflow, Weights & Biases, Kubeflow, or cloud-native services are configured to match your team's skills. Retainers keep systems healthy after launch—retraining schedules, incident response, and quarterly reviews of feature freshness. This is applied AI operations, not shelf software resale.
Typical range: CAD $12,000 – $45,000 setup; CAD $4,500 – $18,000/month retainer
05
Analytics & BI modernisation
We refactor sprawling spreadsheet chains and fragile Looker or Power BI models into governed semantic layers backed by reliable SQL. Dashboards are redesigned around decision workflows—not vanity metrics. Where appropriate we introduce lightweight experimentation frameworks so product teams can test changes without bypassing data governance. Deliverables include style guides, metric definitions, and training sessions for internal analysts who will own the tools day to day.
Typical range: CAD $15,000 – $75,000 per programme
06
AI governance & PIPEDA readiness
We help legal, security, and data teams align on policies for model use, personal information handling, vendor assessments, and documentation expected under PIPEDA and sector guidance. Deliverables may include data processing inventories, impact assessment templates, model cards, and review checkpoints before production promotion. We are consultants, not lawyers—but we translate technical reality into language privacy officers and auditors can work with.
Typical range: CAD $8,000 – $40,000 per assessment cycle
