A small, senior practice. Built for the work.
DaVinci AI is a focused data, analytics and applied-AI practice working with business and public-sector teams across regulated industries.
Founded on a stubborn idea: that data should change decisions.
DaVinci AI was founded by practitioners who’d spent enough years inside large data programs to see the pattern: enormous investment in tooling and headcount, very little change in how decisions actually get made.
The firm is built around the opposite premise. Small teams of senior people. Engagements scoped around a specific decision worth changing. Prototype in weeks; production in a quarter. Honest about what the tools can and can’t do.
Our deepest pattern of work is in environments where the data is sensitive, the oversight is real, and the documentation has to land before the system does, places where being wrong is expensive and being late is worse. The same engineering discipline shows up across financial services, healthcare, infrastructure and the institutional enterprise.
How we choose to work.
The work itself
We are practitioners first. Every engagement is led by the people doing the work, not by partners who pass it down.
Decision over decoration
A dashboard, a model, a workflow exists to change a decision. If it doesn’t, it doesn’t ship.
Production is the deliverable
A prototype proves it’s possible. A production system is what the institution actually runs on. We stay with what we shipped.
Evidence-grade
Every recommendation traces to data, every model to its evaluation, every agent action to its constraint. Assurance is part of the build.
Selectivity over scale
We decline engagements that aren’t set up to succeed. We do not grow on the back of work we wouldn’t stand behind.
Quiet by default
Our clients are referenced when they ask to be. Otherwise we don’t talk about them. Discretion is the default posture.
A practice, not a pyramid.
Every engagement is led by the people doing the work. We staff small, senior teams, typically two to four practitioners across data engineering, analytics and machine learning, rather than rotating in juniors to scale headcount.
That keeps quality consistent, communication direct and the work accountable. For larger programs we partner with a small network of trusted specialists rather than expand the firm.
What we do, and what we deliberately don’t.
Two lists. Where we add real value, and where we’d send you elsewhere. Together, they’re how we keep the practice sharp.
What we do
- Decision-grade analytics and dashboards
- Forecasting, segmentation and decision models
- LLM workflows and document intelligence
- Modern, governed data platforms
How we work
- Small senior teams (2–4 practitioners)
- Discovery sprints in 2–4 weeks
- Build sprints in 8–12 weeks
- Quarterly embedded engagements
Where we work
- Institutional and oversight programs
- Financial services and insurance
- Healthcare operations
- The institutional enterprise
What we don’t do
- Pyramid staffing or junior-only teams
- Multi-year transformation programs
- Direct clinical decision-making
- Investment recommendations or trade execution
Tools we lean on
- dbt, Snowflake, Databricks, BigQuery
- Python, PyTorch, scikit-learn
- Tableau, Power BI, Looker
- Azure OpenAI, AWS Bedrock, Vertex
How we contract
- Mutual NDAs, MSAs, DPAs, BAAs
- Direct and via approved prime relationships where required
- Milestone-based pricing for build sprints
- Client-owned IP with limited reuse license
Frameworks the work runs inside.
Every engagement ships with the evidence an oversight function actually asks for. Model cards, evaluation reports, data lineage, decision provenance, runbooks. Assurance is part of the build, on day one.
- SOC 2 Type II
- ISO 27001 aligned
- HIPAA / HITECH
- GLBA
- SR 11-7 model risk
- NIST AI RMF
- NIST CSF 2.0
- PCI DSS 4.0
- GDPR · CCPA
- EU AI Act readiness
- Customer-tenant deployment
- Evidence-grade documentation
A short history of the firm.
The firm we set out to build, in five movements.
- 2024
Founded
The firm was founded around a stubborn premise: that small teams of senior practitioners, doing the work themselves, would outproduce larger firms running pyramid staffing models. - 2025
First engagements
Initial engagements across financial services and institutional operations. Patterns we still use today, the Diagnostic / Build / Steward engagement model, the decision-first analytics method, the data-product operating model. - 2025
Practice areas formalized
Four disciplines articulated as a single practice: data & analytics, machine learning, applied AI, and data engineering. Reusable patterns moved into a shared internal library. - 2026
Public-facing writing
The first wave of insights published, practice notes drawn from real engagements, stripped of identifying detail and written by the people who shipped the work. - Today
Selective growth
We hire slowly. We turn down engagements that aren’t set up to succeed. We’d rather stay small and do excellent work than grow on the back of work we wouldn’t stand behind.
Questions we hear, answered honestly.
How big is the firm?
Where are you based?
Do you do speaking and panels?
Do you take pro bono work?
Continue with our work.
What we do
A focused practice across data, analytics, machine learning and applied AI.
Explore
How we think
Practice notes, perspectives and reference architectures from the work itself.
Explore
Join the practice
Senior roles for practitioners who want autonomy, ownership and consequential problems.
Explore
Have a workload worth getting right?
If you’re scoping a system someone will rely on, replacing a reporting estate that no longer holds up, or evaluating where AI genuinely belongs in your operations, we’d like to hear about it.