Steve Fusaro
Engagements

A selection of recent client engagements.

Every engagement is scoped, time-boxed, and designed to leave the client team stronger. Filter by category to see the kinds of work that map to your situation.

  • AI & LLM2025

    RAG-powered support copilot

    Series B SaaS

    Designed and shipped a retrieval-augmented support assistant grounded in 40k internal docs, reducing median ticket handle time by 38%.

    • Cut median handle time 38%
    • Sub-2s P95 retrieval latency
    • Eval harness owned by in-house team
    PythonOpenAI / LLM APIsRAG PipelinesVector DBsPostgreSQL
  • Fractional Leadership2024

    Fractional CTO through Series A

    Pre-seed → Series A fintech

    Embedded as fractional CTO for 14 months: built the engineering org from 2 to 11, set technical direction, and led the technical diligence for an $18M Series A.

    • Hired 9 engineers across infra, product, ML
    • Led technical diligence for $18M raise
    • Defined 18-month architecture roadmap
    TypeScriptPythonAWSPostgreSQL
  • Architecture Audit2024

    Scaling audit for a B2B marketplace

    Series B marketplace

    Three-week architecture and reliability audit ahead of a 5x traffic event. Delivered a prioritized 90-day remediation plan adopted in full.

    • Identified 4 critical scaling risks
    • 90-day remediation plan adopted
    • Survived launch with zero P1 incidents
    PostgreSQLAWSKubernetesTerraform
  • Platform & Infra2023

    Monolith-to-services migration

    Healthtech scale-up

    Led a strangler-pattern extraction of three high-churn domains from a 9-year-old Django monolith with zero customer-facing downtime.

    • Three services extracted in 6 months
    • Zero downtime migration
    • P95 latency improved 2.4x
    PythonDjangoPostgreSQLKubernetes
  • Product Engineering2023

    Experimentation platform v1

    Consumer subscription startup

    Built the first version of an in-house A/B testing platform, unblocking the growth team to run 4x more experiments per quarter.

    • 4x experiment velocity
    • Self-serve for non-engineers
    • Owned by growth eng on day one
    TypeScriptReactPostgreSQL
  • AI & LLM2025

    LLM evaluation & observability

    Vertical AI startup

    Set up offline evals, regression suites, and tracing for a multi-agent product. Caught a silent quality regression on the third deploy.

    • Offline eval suite + CI gating
    • Tracing across 7 agent steps
    • Cost down 22% via routing
    PythonOpenAI / LLM APIs

Ready to harden your stack?

Currently accepting strategic advisory and deep-technical projects.

Book a 30-min intro call →