01 / Portfolio · AI TA Engineering & Adoption
$whoami
I architect talent systems for the AI era — building, governing and driving adoption at scale.
Hi, I'm Leo. I build AI systems, talent intelligence and sourcing infrastructure at Canva — the tooling behind 31 recruiters across five global locations.
From signal to placed hire — one operating system, not a pile of tools.
Fourteen years of technical recruiting, now expressed as software, data and models — shipped in production, not theory.
I build the systems, set the guardrails and train the teams — so AI in Talent moves from demo to daily operation.
Finding, assessing and shortlisting are kept separate. Every assessment is tied to a saved profile version and criteria, with source citations, human approval, spend caps and local records.
Talent intelligence platforms, Cortex chat agents, semantic views, agentic sourcing pipelines and bespoke programmatic hiring tools.
PII-redacted prototypes, read-only data access, citation validation, spend caps and human approval before paid or published work.
Training recruiters, writing the playbooks and staying close enough to the work that AI becomes daily practice, not a demo.
Organisational impact first, then how it's executed.
Companies I've helped hire for
Recruiter hires made using talent-search systems I built.
Agent-enabled talent intelligence for hard-to-fill roles.
Compared with an 18% team benchmark.
Based on comparable outreach activity during the same measurement period; methodology available on request.
Live dashboards with recruiter snapshots, automated reporting, and AI agent flows for org-wide training.
Pipeline visibility and queued roles per recruiter — modelling shows who sits under or over capacity, so TA members move where needed without hiring more.
Trained recruiters to build with AI, not just use it — plus the governance layer so they build safely within regulation.
Data sharpens search, search trains the models, the software makes it repeatable.
Copilots that feel like a sharp colleague, not another dashboard — and playbooks that turn “AI in recruiting” into something a team can run on Monday.
Headcount as a supply chain: leading indicators and scenario plans a CFO will actually sign off on.
Market maps, targeting, and engagement for Senior Director+ roles — the searches that cannot be run on volume.
Agentic research pipelines, signal tracking, and a talent OS that makes the whole system repeatable.
Actively building, forecasting, writing and training.
agentic sourcing — software that quietly runs the boring 80% of a search, so recruiters keep the conversations.
building governance layers to use AI in TA safely in an ever-changing regulatory landscape.
playbooks that turn “AI in recruiting” from slogan into something a team can run on Monday.
recruiters on the models — copilots that feel like a sharp colleague, not another dashboard.
If you're scaling a technical organisation — or turning hiring from guesswork into a measurable system — that's the conversation I like having.