2–3 weeks
AI & Architecture Diagnostic
Audit data pipelines, identify LLM/ML feasibility, and deliver an executive execution roadmap for the highest-leverage moves.
- Pipeline & data audit
- LLM/ML feasibility assessment
- Executive execution roadmap
Senior-practitioner consulting
DataSix
Big 4 rigor without the Big 4 bloat.
Boutique data, analytics, and AI/ML consulting where clients work directly with ex-Big 4 and tech-company veterans who can shape strategy and ship the work.
About the principal
Fannie (Fan) Cai
Founder & Principal AI/ML Consultant
Fannie brings an uncommon mix of executive leadership and hands-on delivery: Director of ML & AI Engineering in regulated industries such as finance and healthcare, an ex-Big 4 consultant, and a University of Waterloo lecturer and researcher.
Director of ML & AI EngineeringRegulated industries: finance & healthcareEx-Big 4Waterloo lecturer & researcher
Connect with Fannie on LinkedIn Industry-first edge ML
Real-time predictive models deployed directly across browser and mobile experiences at national scale.
32% revenue lift
Real-time predictive models connected directly to commercial action.
Peer-reviewed AI research
Published in IEEE Access on generative synthetic data and privacy-preserving AI.
Recognized industry voice
University of Waterloo lecturer and keynote speaker at CDAO Canada.
Senior-practitioner model
Clients work directly with ex-Big 4 and tech-company veterans.
Execution with governance
Speed for founders, reliability for enterprise executives.
ROI-first roadmaps
Every engagement is anchored to a measurable business outcome.
Production AI
LLM workflows and predictive models
Executive analytics
Dashboards leaders actually use
Data strategy
Roadmaps tied to business priorities
Governance
Reliable systems with clear ownership
Structured engagements
You know the problem. These engagements solve it.
No discipline buckets to decode — each engagement is scoped, time-boxed, and anchored to an outcome you can point to.
6–8 weeks
Production ML & LLM Sprint
Build and deploy a production-grade predictive model or LLM workflow tied directly to a commercial KPI.
- Production-grade model or workflow
- Deployment with monitoring
- KPI-tied measurement
Retainer
Fractional Head of AI & Data
Strategic leadership, team hiring playbooks, and architectural governance for teams that are scaling fast.
- Strategic leadership on demand
- Team hiring playbooks
- Architectural governance
Proof points
Work that ties technical depth to commercial results.
Type 2 Diabetes Risk Prediction
Time-dependent AUC of 0.87
Five-year diabetes risk prediction models used routine blood biomarkers and longitudinal laboratory data to support earlier risk identification in regulated healthcare.
Sales Funnel Dashboard
32% revenue increase in month one
End-to-end tracking from lead capture to product utilization gave leaders a single operating view of conversion and expansion.
Data Team Structure Playbook
A scaling model for growing teams
Organizational design, role hierarchy, and decision rights helped a scaling company invest in the right data capabilities at the right time.
Research & talks
Published work behind the advisory lens.
Fannie's public research and conference talks reinforce DataSix's focus on responsible, production-minded AI in highly regulated settings.
Research
Routine Blood Biomarkers Predict Five-Year Type 2 Diabetes Risk
Presented at the American Diabetes Association Scientific Sessions, connecting routine lab data to preventive health risk modeling.
Read abstractPublication
Adapted GAN-Based Models for Synthetic Longitudinal Clinical Biomarker Data
Peer-reviewed IEEE Access publication on privacy-preserving synthetic clinical data for regulated healthcare AI.
View paperTalk
Operationalizing ML & AI in Regulated Healthcare Environments
CDAO Canada keynote and panel discussion on building trust, scale, and impact with applied AI.
View talk postVideo
AI in Pharma: Real-World Applications
Future Pharma Marketing Conference talk on hybrid Temporal Convolutional Network and Transformer modeling for healthcare data.
Watch videoTestimonials
What clients say after the work ships.
Fannie is a pleasure to work with and is an exceptional data consultant. She provided oversight to our analysts and helped with the creation of automated SQL scripts and the outlining of a master plan for setting up our data architecture in Azure. Fannie has the ability to understand complex data structures and translate business requirements into efficient, robust queries. I wholeheartedly recommend Fannie as a data consultant. Her dedication to the technical field of Data, coupled with her exceptional communication skills, sets her apart in the field.
Wajdi Ghoussoub
SVP FinancePlooto
Engagement model
Boutique by design, rigorous by background.
DataSix is built for clients who want partner-level thinking in the room and hands-on builders at the keyboard. The model removes layers, meetings, and vague deliverables so the work stays close to business value.
For scale-ups
Rapid diagnostics, implementation sprints, and practical data products that show ROI quickly.
For enterprises
Governed architectures, executive roadmaps, and reliable delivery patterns that withstand scrutiny.
For founders
A direct senior team that can translate ambition into analytics, AI, and data engineering priorities.
For executives
Clear trade-offs, credible sequencing, and decision-ready artifacts without the consulting theatre.
Consultation
Put the highest-leverage move on the calendar.
Choose a time for a focused 30-minute strategy call with a senior practitioner, or send a short note if you prefer to start asynchronously.
30-minute strategy call
Choose a time that works.
Eastern Time · Video call
Calendar not showing?Open Calendly