Geoff Leediker

AI-enabled cloud developer building practical automation tools

Skills

Skills in Practice

Serverless Architectures

Implemented Lambda + API Gateway patterns for event-driven workflows in a conversational RAG pipeline.

Embeddings & Vector Search

Used Titan Embeddings v2 + OpenSearch to enable semantic retrieval across multi-session chat context.

Event Scheduling

Configured EventBridge rules to trigger web scrapers and model-based recap generation on league cadence.

Cloud Security Awareness

Ensured least-privilege IAM roles, secret isolation, and public asset policy hardening on S3 static hosting.

Model Prompt Engineering

Structured prompts for consistency, hallucination mitigation, and domain-specific output formatting.

Monitoring & Observability

Used CloudWatch logs/metrics to diagnose throughput issues and iteration cadence during model testing.

Education & Programs

AWS Cloud Institute • Applied cloud architecture, serverless design, AI platform workflows

Certifications

Graduate, AWS Cloud Institute

What I'm Learning Now

Continuous Growth

Projects

Conversational RAG Pipeline on AWS

Built a serverless Retrieval-Augmented Generation system enabling conversational Q&A over private documents using Titan Text Embeddings v2 and OpenSearch Serverless. Ingestion pipeline handles chunking, metadata tagging, and DLQ retries. Query path emphasizes business-context–aware retrieval and latency control.

Outcome: Reduced time spent searching institutional knowledge
Retrieval Accuracy ↑28% Storage Lifecycle ↓35% p95 Latency 210ms
  • S3 ingestion with EventBridge triggers
  • Lambda for chunking and embedding generation
  • Titan Embeddings v2 for vector representation
  • OpenSearch vector index for semantic retrieval
  • API Gateway inference endpoint
  • Presigned URL upload boundaries
  • Least-privilege IAM roles
  • DLQ for ingestion failures
  • CloudWatch logs for latency and token throughput

Fantasy League Companion (Sleeper + Bedrock)

Built a league companion that ingests fantasy football data from Sleeper, aggregates daily NFL news via EventBridge-triggered Lambda web scrapers, and prompts a Bedrock model to produce weekly recaps, power rankings, and matchup previews. Focused on structured prompt formatting and data freshness.

Outcome: Increased league engagement with timely, data-aware summaries
Data Freshness ↑ Daily Engagement ↑ User Feedback Event-Driven Scheduling
  • EventBridge schedules web scraping Lambdas for NFL news
  • API calls to Sleeper for league statistics and rosters
  • Structured prompt templates for consistent outputs
  • Aggregation pipeline normalizes scoring and rankings
  • Presigned URLs isolate upload surfaces
  • CloudWatch logs trace execution and failures

Currently extending to surface trend deltas and injury implications.

Static Resume Site on AWS (S3 + CloudFront)

Designed and deployed a fast, responsive resume website hosted on S3 with CloudFront caching. Implemented correct MIME type configurations, public-read bucket policies, and accessibility affordances including skip links, reduced-motion support, keyboard focus rings, and semantic landmarks. Added dark/light theme persistence via localStorage and system preference detection.

Outcome: Improved recruiter scan efficiency and site performance
p95 Load 120ms (CDN) Zero Compute Cost Accessible Interactions
  • S3 static hosting with correct bucket policy boundaries
  • CloudFront distribution with edge caching
  • Dark/light mode theme persistence
  • Reduced-motion preference support
  • Skip-to-content and semantic aria landmarks
  • Print stylesheet for recruiter workflows

Transitioning into cloud-based ML platform support, leveraging prior experience in ambiguity resolution, inventory accuracy, and structured onboarding.

Professional Experience

Area Sales Manager

Performance Foodservice • 2022–2024 • Regional

  • Prospected and onboarded new accounts while managing a large existing portfolio across a wide geographic territory.
  • Drove ~40% growth in total case sales during summer 2023 by identifying operational bottlenecks and improving ordering cadence.
  • Exceeded 70% in total brand sales, contributing to category adoption and supply alignment.
  • Reduced customer churn risk by translating ambiguous product requirements into repeatable ordering workflows.
  • Won two separate performance promotions based on measurable growth and account expansion.

Fulfillment Operations Manager / FSL Specialist

Enjoy Technology / Asurion • 2021–2022 • On-site

  • Managed inventory accuracy and order fulfillment while coordinating with leadership on experimental operational procedures.
  • Oversaw Apple partnership onboarding, including asset registration, SOP interpretation, and training new staff on partner workflows.
  • Served as test market for new operational processes prior to nationwide rollout, providing structured feedback on scalability and failure modes.
  • Trained new hires on complex SOPs, improving consistency and reducing operational defects across distributed teams.
  • Recognized by national leadership as one of four specialists to exceed KPIs in consecutive quarters.

Stay-At-Home Parent

2024–Present • Remote

  • Maintained household logistics while balancing competing time-sensitive priorities.
  • Developed task batching, scheduling discipline, and lightweight documentation habits.
  • Used flexible availability windows for independent study of cloud services, AI workflows, and platform fundamentals.

About

I build practical cloud and ML tooling that improves delivery speed and reduces cognitive overhead for teams. I value simplicity, observability, and continuous refinement.

Contact

Seeking opportunities developing cloud infrastructure for ML workloads.

Email Me