The Operator
Infrastructure Leader
Who Ships

Specialized in big data analytics—terabytes of healthcare data, machine learning, finding patterns in massive datasets.

Began as a scrum master (which I'll admit isn't real project management). Started coding in DevOps to help the team when engineers couldn't fix issues. Taught myself software engineering to solve production problems.

Grew to VP managing engineers across multiple continents. Owned IaC across three major clouds. Passed Australian IRAP security assessment. Became the "incident response commander"—the person called when production goes down to lead the warroom and fix it fast.

Built AxilPy - HCP analytics platform (Next.js/TypeScript, entire codebase). Analyzes 100k+ providers for pharmaceutical targeting. BigQuery backend, real-time dashboards, weighted scoring engine.

Built 3 AI Agents (PyTorch/Python) for rare disease detection using 7 years of American claims data:

Agent 1 (Agentic-to-Agentic): Framework learning optimal ML params → +20% F1 improvement
Agent 2 (Neural Features): Self-learning network reading disease papers, formulating experiments → +20% additional boost via ensemble
Agent 3 (Sequences): No manual features—raw claims data only. Beta validation.

Managed NIH grants, led through healthcare market cycles and long clinical validation timelines.

Building breakthrough technology is exciting, but the best leaders know when to ship pragmatic solutions that keep the business moving. Innovation is valuable when it serves the mission—not when it becomes the mission itself.

Building AI agents taught me they're workflow optimizers, not magic. The technology evolves incredibly fast—code I wrote in May looks radically different from November because of how quickly core models improve. It's "living code" that requires continuous learning and adaptation.

Week 1-2: Listening Tour
• 1:1s with every direct report and key stakeholders (App, Data, Security, Product)
• Shadow on-call rotation to see incident reality
• Inventory critical services: "What breaks the firm if it goes down?"
• Review last 6 months of incidents (P0/P1) for patterns

Week 3: Risk Assessment
• Top 10 reliability risks (single points of failure, tech debt, runbook gaps)
• Security gaps (access controls, patching cadence, secrets management)
• Cost hotspots (untagged resources, orphaned infra, overprovisioned instances)
• Data pipeline health (SLA violations, data quality issues, freshness)

Week 4: Foundation Setup
• Define 5-7 core SLOs for "tier 1" services
• Incident severity model + escalation paths
• Service catalog: owners, dependencies, SLAs, runbooks
• Establish on-call expectations and support tiers

Quick Wins Delivered:
• Fix 2-3 "easy but annoying" issues (alert noise, broken dashboards, stale docs)
• Ship 30-day assessment deck to leadership

Observability Foundation:
• Standardize logging (structured logs, retention policies)
• Metrics baseline (golden signals: latency, traffic, errors, saturation)
• Distributed tracing for top 5 critical paths
• Dashboards for each service tier (SLO burn rate, error budgets)

Incident Response Discipline:
• Alert hygiene: reduce noise by 50%, ensure every alert is actionable
• Runbook template + populate for top 10 services
• Incident command structure (IC, Comms, Ops, Eng leads)
• Weekly incident reviews → engineering work (not blame)

Change Management:
• CI/CD pipeline gates (testing, security scans, manual approval for prod)
• Infrastructure as Code standards (Terraform modules, state management)
• Deployment windows and freeze calendars

FinOps Program:
• Tagging strategy (cost center, environment, owner)
• Budget alerts and guardrails per team/project
• Monthly cost review with engineering leads
• Identify 3-5 cost optimization opportunities (rightsizing, reserved instances, cleanup)

Infrastructure Scaling:
• Golden-path Terraform modules (VPC, app stack, data pipeline templates)
• Environment vending (devs can spin up staging/dev in 10 minutes)
• Integration playbook: SaaS ↔ internal ↔ vendors (identity, network, observability, data)
• Integration playbook with Day 1/30/90 gates and metrics

AI Enablement Program:
• Launch enterprise AI tool adoption (Cursor, ChatGPT, n8n, Zapier) for every department
• Force multiplier: AI makes the org work faster and look bigger by spending less on headcount
• Persona-based workshops (GTM, Eng, Ops, CS) moving teams curiosity → confident usage
• "Happy paths" library: reusable templates for high-ROI workflows
• Self-serve enablement hub (runbooks, FAQs, video walkthroughs, what's new)
• Track adoption metrics: WAU/MAU, template reuse, time saved, ticket deflection

Team Scaling & Culture:
• Hiring plan: SRE nucleus (2-3 hires) + Cloud Platform Leads (1-2 senior)
• Leverage my network for fast, quality hires
• Culture > Talent: Hire for values alignment first, skills second
• Internal growth: promote 1-2 high performers into stretch roles
• Champions network for AI fluency and best practices sharing

Deliverables by Day 90:
• SLO dashboards live for all tier-1 services
• 3 "golden path" modules deployed to production
• Cost optimization delivering 15-20% reduction
• First AI enablement cohort graduated (50+ users trained)
• Team hiring pipeline filled with 5+ qualified candidates