Profile
I advise enterprises on applying AI platforms, systems and governance frameworks—designing them with their teams, selling and overseeing the implementation, and building the technical teams that deliver them.
Over twelve years my work has moved from forensic data investigations to enterprise AI. At Dataiku I was the first sales engineer in London, then made the first hires in new regions, seeding teams that have since grown to 40+ engineers. I lead as a player-coach: hands-on in the most complex engagements while developing the people around me. Today I own go-to-market for Dataiku’s AI governance product, working with CDOs, CISOs, risk and compliance teams across EMEA and APAC. Outside work I run my own AI stack: a local inference server and agents distributed across VMs in a homelab I’ve built myself. I also sit on the board of Alumni Europae ASBL.
01 / Selected work
A few engagements and builds that show how I work.
Designed Dataiku’s EU AI Act compliance and model risk management offerings from first concept to a defined product and reference architecture, then took them to market and equipped sales teams across EMEA and APAC with the architectures and collateral to sell them. The offering influenced over £1m in ARR in its first year, and field feedback from clients shaped features the product team later shipped.
Joined as Dataiku’s first sales engineer in London and defined the role. As Director, I made the first hires in new regions across the UK & Ireland, the Nordics and the Middle East, seeding teams that grew into 40+ engineers. I hired seven people directly and led up to eight direct reports; four of those hires have since stepped up into management or principal roles. I stayed hands-on throughout, personally leading what was then Dataiku’s largest EMEA deal as we reached 121% of target in FY2020.
End-to-end engagements (discovery, workshops, proofs of concept, implementation) translating the EU AI Act, GDPR, DORA and NIS2 into working controls for clients including a European Ministry of Defence, European-headquartered global banks and a Middle Eastern central bank.
Managed and helped build a transaction-review and compliance platform, with full audit trail and RFI workflow, that was central to a New York DFS-mandated monitorship of a large European bank. I presented the work to the DFS in 2014.
02 / Experience
Twelve years across investigations, sales engineering and AI governance. The full detail is in my CV.
Dataiku
Owns go-to-market for Dataiku’s AI governance product: built its EU AI Act and model risk management offerings, leads engagements with banks, insurers, defence and public-sector clients, and equips sales teams across EMEA & APAC to sell them.
Dataiku
Made the first sales engineering hires in new regions across the UK & Ireland, the Nordics and the Middle East, seeding teams that grew to 40+ engineers. Hired seven and led up to eight direct reports; 121% of target in FY2020.
Dataiku
First sales engineer in the region: defined the role and its success metrics, and landed flagship clients across financial services, healthcare, consumer goods and telco.
Deloitte
Managed analytics teams across three workstreams in a global investigation of a German automotive manufacturer, working with petabyte-scale data for legal proceedings.
Berkeley Research Group
Developed the analytics engine behind BRG’s DRIVE™ investigations platform and the controls protecting the sensitive client data it held.
FTI Consulting
Managed and helped build a transaction-review and compliance tool for a US regulator-mandated monitorship of a large European bank, linking SWIFT and transaction data across Europe and the Americas.
03 / Talks
On governing AI in practice, from analytics to generative AI and agents.
Keynote · AI Week, Frankfurt
What changes for governance when AI systems start acting on an organisation’s behalf.
Web series · Dataiku
How organisations can get full visibility over their AI systems, from development to deployment, using Dataiku Govern.
Read the recapKeynote · Big Data London
What generative AI changes for governance programmes, and what stays the same.
Keynote · Big Data London
Governance as a way to scale AI, rather than a brake on it.
Keynote · SIGDSA (virtual)
Keynote for the Special Interest Group on Decision Support and Analytics.
Talk · EGG On Air
Using the underlying data to move beyond headline statistics and understand what the gender pay gap actually tells us.
Watch the talk04 / Writing
Writing on the evolving rules around AI, data and institutional accountability.
Article · Dataiku
AI governance and the cost of (non)compliance.
Article · Dataiku
How regulation is adapting to rapid AI development—and what that means for analytics leaders.
Read article (archived)Article · Thomson Reuters
Where the EU’s proposed cybersecurity directive fell short for the organisations it aimed to protect.
Article · Thomson Reuters
Managing growing volumes of corporate data under rising regulatory expectations.
05 / Homelab & code
What I advise on, I also build. I’ve been building computers since I was 14 and have run my own infrastructure since 2004.
Multi-agent systems
Agents run across several virtual machines and harnesses on Linux and macOS. Much of the work has been tuning them until they hold a coherent multi-agent flow on complex, multi-step projects.
Production project · Private repository
A production-grade member portal for Alumni Europae ASBL, where I sit on the board. Because it holds a sensitive member database, it is built to production standards by the multi-agent development team I’ve designed and tuned across my self-hosted VMs.
Local inference
A dedicated server for running models locally, serving the agent stack so that experiments and data stay in-house.
Hardware
Two self-built servers alongside Intel NUCs and Raspberry Pis across multiple sites, running Proxmox, unRAID and Portainer.
Network & security
OPNsense routing with VLAN segmentation, a Tailscale and Pangolin zero-trust mesh between sites, and CrowdSec intrusion prevention across 50+ self-hosted open-source services.
GitHub · @sbmontgomery5
Public repositories, experiments and open-source contributions.