KBR
Defense analyticsTechnical owner of a defense analytics platform
I own a regulated analytics platform, optimize dashboards over 5M+ records, and deliver demos to government product owners.
Kasey Schoeff
Full-Stack Engineer | AI demos, prototypes, and customer-facing systems
Rapid prototypes backed by real systems.
Built across 5M+ production records, 95+ shipped endpoints, and 26-agent orchestration systems. I scope use cases quickly, ship working AI prototypes, and turn complex technical behavior into demos customers and stakeholders can actually follow.
Contact
kasey.a.schoeff@gmail.comLinkedIn profile
Experience
Work contexts that shape how I build and demo.
Defense analytics, startup product engineering, founder-led AI implementations, and a teaching foundation that keeps technical systems legible to different audiences.
Cracked Gaming
Startup productFounding engineer on a zero-to-production startup build
I ship payments, real-time systems, Twitch integrations, creator tooling, and operator tooling across 95+ API endpoints.
Founder-led AI product work
Applied AIFounder-led AI implementations
I build OpenAI-powered prototypes across evaluation workflows, semantic matching, job intelligence, and multi-agent orchestration.
STEM tutoring
TeachingTeaching and technical communication
Tutoring foundation across calculus, physics, chemistry, statistics, and algebra for hundreds of AP students.
Selected work
Projects framed for customer-facing AI work.
Each case study focuses on the problem, what was built, and why it mattered.
Trust-sensitive evaluation workflow
Built an evidence-backed credentialing workflow for trust-sensitive AI evaluation.
Made model judgments legible with grounded evidence, review stages, and a verification layer.
EvaluationHuman reviewVerifiable outputs
Why it matters
Evaluation, review loops, and verification make model decisions easier to trust and deploy.
What shipped
Shipped a four-stage workflow for submission review, evidence grounding, adversarial checks, and credential verification.
Key patterns
Evidence groundingReview workflowVerification layer
Core stack
Next.jsTypeScriptSupabaseOpenAIPlaywrightVitest
What stands out
- Explains evaluation and trust in plain language.
- Pairs human review with model judgment.
- Transfers well to regulated or high-stakes workflows.
Why OpenAI
Why this profile maps to OpenAI's Demo Experience Engineer role.
This is the overlap I already work in: customer-facing demos, rapid prototypes, and OpenAI API implementations that have to feel useful and credible to the audience in front of them. The strongest evidence is the mix of government stakeholder demos, startup product delivery, and applied OpenAI work across evaluation, retrieval, matching, and orchestration.
Customer-facing demos
Government stakeholder demos and startup operator presentations with an emphasis on clarity and business value.
Rapid prototyping
Prototype development across evaluation pipelines, market intelligence, matching systems, and internal AI tooling.
OpenAI API experience
OpenAI API work across structured outputs, embeddings, evaluation, retrieval, and multi-step workflow orchestration.
Trust-sensitive systems
Regulated environments, review loops, verification, moderation, and reliability constraints relevant to enterprise AI.
Data and retrieval
Real systems behind the demos: dashboards, job intelligence, semantic matching, retrieval flows, and operational tooling.
Technical translation
Model behavior, product tradeoffs, and architecture decisions translated into demonstrations non-technical audiences can follow.