Know your customers. Grow faster.
I use machine learning and AI to help companies — and the consulting firms that serve them — turn product usage, CRM, digital and transaction data into decisions about who to target, what to build and where to invest. Fifteen years in data science and marketing analytics, former BCG and Accenture practice lead, and the rare combination of someone who leads the engagement with your stakeholders and builds the models.
Transforming consumer data into measurable growth frameworks.
Each framework pairs a business question with the machine-learning methods that answer it — so the output is a decision your marketing, product and sales teams can act on, not a report.
Audiences that matter
Uncover the customers and accounts that matter most with data-driven precision — who buys, who's about to leave, and who's worth the next dollar.
- Segmentation & persona modeling
- Propensity, churn & retention prediction
- Purchase-influencer and ICP identification
- Brand health across a portfolio
What drives conversion
Identify what truly drives conversion — and where to invest for maximum impact across channels, campaigns and product pairings.
- Marketing mix modeling & attribution
- CAC / LTV & spend optimization
- Campaign diagnostics & incrementality
- Product-pairing & basket analysis
Adoption and time to value
Predict adoption, remove friction and guide users to value faster — for new features, new products and new AI capabilities.
- Feature- and AI-adoption readiness scoring
- Usage-based segmentation & activation analytics
- Onboarding friction & drop-off diagnosis
- GTM, ICP refinement & buyer-journey insight
The models behind the insights — built, deployed and monitored.
Every framework above is backed by working code. These are the deliverables clients come back for.
Product relationships from single-item transactions
A proprietary association-rules model that reveals hidden product relationships where conventional basket analysis finds nothing — proven on a global apparel maker's e-commerce catalog.
AI-integrated dashboards
Interactive dashboards are only the beginning: prediction, optimization and simulation models embedded in Tableau and Power BI, so business users run scenarios where they already look at their numbers.
Predictive, vision, NLP and LLM models in production
Account- and user-level scoring, computer vision on product imagery, NLP on customer text, and LLM-powered workflows for marketing, sales and service — deployed where your stack lives and monitored after launch.
Most teams hire a lead who can't build, or a builder who can't lead. I do both on the same engagement.
I own the room and the roadmap
- Scope the business question with executives and marketing, sales and product stakeholders
- Run the workstream: plan, cadence, steering-committee readouts, change management
- Lead and mentor cross-functional teams of data scientists and consultants
- Translate model output into the decision, the slide and the operating change
I write the code that ships
- Data engineering and feature creation in Python, SQL, PySpark, Databricks and Snowflake
- Model development, from classical ML to computer vision, NLP and LLM workflows
- Deployment, monitoring and dashboard / KPI integration
- Documented, handed-over code your team or the client's can own
Two doors. Same senior practitioner behind both.
A senior data scientist you can put in front of your CMO
For marketing, growth, product and analytics teams that need consulting-grade insight without the consulting-firm overhead — and someone who will still be there to build it.
- Fractional analytics lead — a retainer for ongoing insight, modeling and team direction
- Fixed-scope project — a segmentation, MMM, churn model or dashboard with a clear handoff
- Embedded practitioner — hourly or weekly inside your team while you hire or scale
Practice-lead-level capacity for pursuits and delivery
I've sat on your side of the table at BCG and Accenture. I know what a proposal needs, what a deliverable has to look like, and how to run a client workstream under your brand.
- Pursuit support — approach, demo assets and proof-of-value builds that win the work
- Client-facing workstream lead — run the data-science track and the stakeholder relationship
- Senior modeler — hands-on delivery alongside your team, on your paper, under NDA
Insights that changed a decision. Models that shipped.
Product-pairing analysis to lift e-commerce revenue
A proprietary association-rules model that finds product relationships even in single-item transactions, used by a global footwear and apparel maker to drive attach rates online.
Brand health and consumer retention across a portfolio
A machine-learning brand-health model measuring performance, retention and cross-brand engagement — the evidence base behind a portfolio's marketing and growth decisions.
Predicting which organizations will adopt AI-driven features
Account- and user-level readiness scores for new AI features, so marketing targets the organizations most likely to move and the client tracks adoption as a measurable metric.
Detecting emerging color trends from product images
Convolutional networks and clustering that turn a product catalog's imagery into a structured color-trend signal merchandising and design teams can act on.
Scoped like a consultant. Delivered like an engineer.
Scope
A short discovery with your stakeholders to pin the business question, the data that exists, and what "done" looks like. You get a one-page plan with a fixed price or a weekly rate.
Build
Data engineering, features, modeling and a working interface — with weekly readouts so executives see decisions forming, not status reports.
Hand off
Documented code, monitoring in place, and a walkthrough for whoever owns it next — your team or the client's. Optional retainer for iteration.
Fifteen years helping organizations understand their customers.
I'm Steven Baez. I started in model development at Dell and Liberty Mutual, built consumer-insight and predictive models in BCG's advanced analytics practice, then led Accenture's Augmented Insights practice for Fortune 500 clients — moving marketing and sales teams from BI dashboards to machine-learning-driven decisions and managing the data scientists and consultants who delivered them. Today I run an independent practice doing that work for companies directly and for the firms that serve them.
Have a customer question, a dataset, or a proposal due?
Thirty minutes is enough to tell you whether I'm the right fit and roughly what it would take. No pitch.