Nicholas Hopper
Nicholas (Nick) Hopper is an AI engineer at APX Stream who builds agent-driven software delivery systems. He was previously a senior software engineer at ForeFlight and iHeartMedia and holds a computer science degree from Baylor University.
Nicholas Hopper is a third-generation software engineer who has shipped production systems since 2017. His current work is applied AI: multi-agent planners, retrieval pipelines, evaluation suites and model serving for a financial software product at APX Stream.
Before APX Stream, Hopper spent four years at ForeFlight, where he built the real-time Air Traffic feature and led the modernization of a device sync service that processes over 1 billion messages per month. Earlier roles at Intuit, Verde Mining, iHeartMedia and USAA covered real-time data pipelines, optimization software and Apache Kafka migrations.
Hopper also runs his own working systems on AI coding agents, in which several agents hand work to each other through schema-checked files. This site records what he has built and what he has learned from running agents in real delivery.
Experience
- June 2026 to present
AI Engineer, APX Stream
Nicholas Hopper is the AI engineer on APX Respond, an asset management SaaS product that automates due diligence questionnaire (DDQ) and request for proposal (RFP) responses with large language models.
- March 2022 to 2026
Senior Software Engineer, ForeFlight
Nicholas Hopper was a senior software engineer at ForeFlight from March 2022 to 2026, where he built the real-time Air Traffic feature and led the modernization of device-to-device sync serving millions of pilots.
- 2022 to 2024
AI Engineer, Plotr
Nicholas Hopper was a contract AI engineer for Plotr from 2022 to 2024, building a Shopify plugin that gives ecommerce stores an AI sales assistant grounded in their own site.
- 2022
Software Engineer, Intuit
Nicholas Hopper was a contract software engineer at Intuit for nine months in 2022, brought in to resolve Apache Kafka and Apache Camel problems in the data pipelines behind its marketing integrations.
- 2021 to 2022
Software Engineer and Architect, Verde Mining
Nicholas Hopper was the sole engineer at Verde Mining, a Bitcoin mining operation, from 2021 to 2022, responsible for all software and cloud infrastructure.
- November 2018 to December 2021
Senior Software Engineer, iHeartMedia
Nicholas Hopper was a senior software engineer at iHeartMedia from November 2018 to December 2021, integrating the data science team's mathematical solvers into sales software.
- September 2017 to November 2018
Software Developer and Integrator II, USAA
Nicholas Hopper was a software developer and integrator at USAA from September 2017 to November 2018, working on a 15-year-old voice platform.
Projects
- 2026 to present
Floating Texas
Floating Texas is a map-first web app built by Nicholas Hopper that shows current river conditions and access points across Texas.
- 2026 to present
Megawatt Maintenance
Megawatt Maintenance is a marketing site built by Nicholas Hopper for an industrial generator service company covering CAT, Waukesha and John Deere fleets.
- 2026 to present
Memector
Memector is a meme-template search built by Nicholas Hopper that fuses up to five ranking channels with Reciprocal Rank Fusion in pgvector.
- 2026 to present
VDE
VDE is an Electron desktop app built by Nicholas Hopper that reads a git repository's history and reports fourteen structural and historical health signals for agent-written codebases.
- 2026 to present
Vibecode City
Vibecode City is a persistent multiplayer Three.js world started by Nicholas Hopper, in which each player's AI coding agent builds and publishes their house through an installable agent skill.
- 2025 to present
ERentals Exclusive
ERentals Exclusive is a luxury rental marketplace for which Nicholas Hopper built the first version of the cross-platform app and the technical SEO, as client work.
Skills
- Agents and retrieval
- Agent harnesses, Agent security, Agent Skills, Computer use, Document parsing, Git worktrees, Knowledge graphs, LangChain, LangGraph, LLM-as-judge, Model Context Protocol, Prompt versioning, Reciprocal Rank Fusion, Reranking, Retrieval-augmented generation
- Models
- Amazon Bedrock, Claude Code, Codex, Fine-tuning, Grok, LiteLLM, LLM evaluation, OpenAI API, Self-hosted inference
- Languages
- Bash, Go, Java, Python, SQL, TypeScript
- Frameworks
- Apache Camel, Astro, D3, Electron, Expo, FastAPI, Next.js, Node.js, React, React Native, shadcn/ui, Spring, Storybook, Tailwind CSS, TanStack Query, Three.js, Vite, WebSockets, Zod
- Cloud and infrastructure
- Amazon Web Services, AWS CDK, Azure Pipelines, Bicep, Cloudflare Turnstile, Docker, Docker Compose, Fly.io, GitHub Actions, Google Cloud, Kubernetes, Microsoft Azure, Terraform, Vercel
- Data
- Amazon Aurora, Amazon DynamoDB, Apache Kafka, Braze, Elasticsearch, JSON Schema, Mapbox, Office Open XML, OpenAPI, pgvector, Pinecone, PostgreSQL, SQLite, Stripe, Supabase, tree-sitter
- Delivery and observability
- GitHub, GitLab, Grafana, Jenkins, OpenTelemetry, Playwright, Prometheus, Puppeteer, pytest, Splunk, uv, Vitest
- Practices
- Architecture decision records, Distributed systems, Programmatic SEO, System design, Technical leadership
Education
- 2018
Graduate courses in Computer Science, University of Texas at San Antonio
Nicholas Hopper took graduate computer science courses at the University of Texas at San Antonio in 2018, focused on cryptography and blockchain technologies.
- 2013 to 2017
Bachelor of Science in Computer Science, Baylor University
Nicholas Hopper earned a Bachelor of Science in Computer Science from Baylor University in 2017, with focuses in earth sciences and software engineering.
Awards
- 2025
Innovation Award, ForeFlight Hackathon
Nicholas Hopper won the Innovation Award at the 2025 ForeFlight Hackathon for an AI-driven interface to the ForeFlight iOS app, which put a 600-page reference PDF in a RAG database behind a text-to-action interface.
Field notes
Agent skill
An agent skill is a packaged set of instructions, usually a SKILL.md file, that an AI coding agent loads to carry out a specific kind of task. Nicholas Hopper writes skills as the operating procedures of his systems and tests them like code.
Agent slop
Agent slop is the low-value output that AI coding agents leave in a repository: duplicated code, redundant documents and commits too large to review. Nicholas Hopper measured it in a real repository and found that it is structural and historical, and invisible to a linter.
Block instead of guess
Blocking is an agent's explicit, recorded stop when it cannot proceed without a person. Nicholas Hopper gives his agents a block command and tells them to use it for sign-ins, quotas and judgment calls instead of guessing.
Burn-down list
A burn-down list is a file of known violations of a rule that is allowed to shrink but never to grow. Nicholas Hopper uses one to enforce a new lint rule in continuous integration from the first day, without first fixing every existing violation.
Diff-scoped gate
A diff-scoped gate is a set of pre-merge checks chosen according to which files a change touches. Nicholas Hopper built one in which a script, not the agent, decides the scope, and in which the scope can be widened by judgment but never narrowed.
Fail closed
Failing closed means that when a component cannot do its job, the system refuses to proceed. Nicholas Hopper applies it at every boundary with a language model, so that a model outage or an unparseable response produces a refusal and never a silent pass.
File handoff
A file handoff is a transfer of work between two AI agents in which the only thing passed is a file that both sides can validate. Nicholas Hopper uses file handoffs so that agents on different harnesses can share a workflow without sharing a conversation.
Git worktree delivery
Git worktree delivery is the process of landing work from several parallel git worktrees, each used by a different AI agent session, onto one main branch. Nicholas Hopper wrote a delivery procedure that proves no agent's commit was lost.
Golden set
A golden set is a fixed collection of inputs with known correct outputs, used to score a model or tool before and after a change. Nicholas Hopper builds a golden set per tool and gates prompt and model changes on the scores.
Green
Green is the state of a check or pipeline that reports success. Nicholas Hopper defines in writing what does not count as green, because a pipeline run by AI coding agents can report success without having tested anything.
Human approval gate
A human approval gate is a point in an agent workflow where the agent must stop and a person must act before the work continues. Nicholas Hopper places the gate at the one irreversible action in a workflow and lets agents do everything before it unattended.
Reciprocal Rank Fusion
Reciprocal Rank Fusion (RRF) is a method for combining several ranked result lists into one by scoring each result on its position in each list. Nicholas Hopper used it in Memector to fuse up to five search channels, and measured where it fails.
Run directory as memory
A run directory is a folder that holds every input, output and decision of one execution of an agent workflow. Nicholas Hopper treats the run directory as the workflow's only memory, in place of chat history or the state of any external tool.
Writing
Files, not chat: how three agents hand work to each other
Nicholas Hopper runs a weekly workflow in which Grok, Claude Code and Codex each own stages and pass work through schema-validated files. This article describes the rules that keep three agents and one person from overwriting each other.
What green means when agents ship to main
When AI coding agents deliver code to a main branch, a passing pipeline is only as good as its definition of passing. Nicholas Hopper has found pipelines that reported success without testing anything, and wrote down what he no longer accepts as green.
When the measuring instrument is wrong
Verifying the work of AI coding agents depends on tests, evals and checkers that can themselves be wrong. Nicholas Hopper has recorded cases where the instrument, not the product, was at fault, and one where an agent changed the instrument to pass.
Outside work
Art and graphic design
Nicholas Hopper was a state finalist artist in high school, where he began with a focus on graphic design.
Brazilian jiu-jitsu and strength training
Nicholas Hopper trains Brazilian jiu-jitsu, in which he holds a blue belt, and strength training.
Downhill skateboarding
Nicholas Hopper won the Gnarathon, the largest downhill skateboarding race in Texas, three times in a row.
Mountaineering
Nicholas Hopper is a mountaineer who has summited Cotopaxi twice, Chimborazo and Mont Blanc.