8000 kchia (Hou C.) · GitHub
[go: up one dir, main page]

Skip to content
View kchia's full-sized avatar

Highlights

  • Pro

Organizations

@Boundless-Avocado @CulturedCheese

Block or report kchia

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
8000
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kchia/README.md

Hi, I'm Hou

I'm a senior forward-deployed engineer with 10+ years building and operating software, now focused on healthcare automation and applied AI.

I work directly with users to understand their workflows, define what is worth automating, and own delivery from discovery through rollout and evaluation. My healthcare background spans quality measurement, patient scheduling, and operational software.

My public AI projects explore retrieval, agent workflows, and evaluation: how to make a system useful, understand its failures, and decide when it should return a task to a person.

LinkedIn · Email · Prompt Deploy

Selected projects

Screenshot → requirements review → React/TypeScript component. A full-stack AI project combining visual analysis, human review of proposed requirements, and hybrid keyword/vector retrieval over shadcn/ui patterns.

The evaluation harness measures stages of the generation pipeline. Its documented scope includes simulated human approval and exclusions from the application workflow.

Video walkthrough · Retrieval code · Evaluation code · Evaluation scope

Reference implementations for resilience, cost control, observability, and evaluation. Python and TypeScript examples covering retries, fallbacks, structured-output validation, tracing, and evaluation harnesses.

Includes documented failure modes, mock-provider overhead benchmarks, and modeled costs.

Evaluation harness · Structured-output validation

Architecture studies starting with a user problem and explicit constraints. Reviews examine data readiness, evaluation, failure modes, rollout, and cost across healthcare, customer support, and other domains.

Background

Earlier engineering work includes healthcare scheduling, payments, and consumer products at Zocdoc, Allergan Aesthetics (AbbVie), and StyleSeat. Before software engineering, I worked on healthcare policy research and quality measurement.

I've taught 500+ students through coding bootcamps and enterprise workshops, and advised teams on architecture and AI product scope.

Princeton University — B.A., cum laude; Senior Thesis Award recipient.

Writing and videos

I publish Prompt Deploy, where I write about workflow design, evaluation, and reliability in AI systems.

Short technical breakdowns: YouTube.

Core tools

Application engineering: Python, TypeScript, SQL, React, Next.js, Node.js, FastAPI, PostgreSQL, SQL Server, Docker, AWS.

AI project work: LLM APIs, retrieval-augmented generation (RAG), LangGraph, Qdrant, evaluation, and tracing.

I'm interested in Applied AI and forward-deployed engineering roles with hands-on ownership from discovery through deployment and iteration. Get in touch.

Pinned Loading

  1. ai-system-design-notes ai-system-design-notes Public

    AI system design reviews — production constraints, failure modes, cost models, and honest trade-offs.

    1

  2. production-llm-patterns production-llm-patterns Public

    Framework-agnostic production patterns for LLM systems — resilience, cost control, observability, testing, safety, orchestration. Dual TypeScript/Python implementations with benchmarks and cost ana…

    TypeScript 1

  3. component-forge component-forge Public

    Multi-agent RAG pipeline transforming UI designs into production-ready React components with design-system integrity.

    Python

  4. langgraph-research-agent langgraph-research-agent Public

    An implementation of cyclic agentic workflows using LangGraph. Demonstrates precise state management and iterative reasoning loops for complex, open-ended research tasks.

    TypeScript

  5. ai-engineering-starter ai-engineering-starter Public template

    A comprehensive 'Day One' template for AI products. Integrates TypeScript/Python with observability, auth, and CI/CD, providing the operational scaffolding needed for reliable LLM applications.

    Makefile

  6. gaming-research-agent-ai gaming-research-agent-ai Public

    A production-oriented research agent utilizing RAG and semantic search to navigate industry-specific data. Features an agentic workflow designed for verifiable outputs and structured information re…

    Jupyter Notebook 2 2

0