Skip to content
View erishen's full-sized avatar

Block or report erishen

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. 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
erishen/README.md

Hi, I'm Lei Sun

AI engineering / platform engineering / senior full-stack engineer focused on making AI useful in real engineering workflows.

I currently work on FinTech engineering at PayPal, with recent focus on AI-assisted payment migration validation, multi-stack e-commerce test environments, Docker / Makefile workflows, Playwright E2E validation, and evidence-driven engineering loops.

I care less about one-off AI demos and more about systems that are runnable, reproducible, testable, and explainable.

πŸ” Open to opportunities


Current Focus

  • AI-assisted engineering workflows for real software systems
  • Agent workflow design with planning, execution, independent evaluation, and evidence collection
  • Multi-language payment validation environments with Docker, Makefile, and Playwright
  • Developer tools, CLI workflows, prompt systems, memory systems, and MCP-based code context
  • Full-stack systems with React, Next.js, Node.js, TypeScript, Python, FastAPI, Redis, and APIs

πŸ“ Shanghai, China

🌐 https://erishen.cn


Featured Projects

An AutoGen-based exploration of Planner / Specialist / Evaluator agent workflows, focusing on task decomposition, execution, independent review, and evidence-driven iteration.

Personal experiments around agent design, workflow orchestration, tool usage, and practical AI execution patterns.

An OpenAI Function Calling-compatible tool registry and CLI for exposing local tools through standardized schemas and agent-friendly interfaces.

A practical LLM toolkit built with LangChain and LiteLLM, supporting chat, RAG document Q&A, hybrid retrieval, streaming APIs, evaluation, and cost-aware usage patterns.

Demo: https://langchain-llm-toolkit.onrender.com/docs

A modern AI chat application with streaming responses, RAG document retrieval, Markdown rendering, theme switching, and practical conversation workflows.

Demo: https://chat.erishen.cn

A personal AI chat application focused on clean interaction, local-first workflows, streaming responses, and lightweight AI product experience.

A full-stack scaffolding and code generation CLI for Next.js, Express, GraphQL, MySQL, TypeScript, CRUD modules, and project initialization.

NPM: https://www.npmjs.com/package/nsgm-cli

A self-hosted shadcn/ui component registry with Storybook documentation, exploring internal component distribution and frontend platform patterns.


Engineering Themes

AI Engineering
Agent workflows, tool calling, MCP, RAG, evaluation loops, prompt systems, memory systems, SSE execution logs.

Platform Engineering
Docker, Makefile, multi-stack environments, reproducible workflows, local tooling, developer productivity.

Full-Stack Systems
React, Next.js, TypeScript, Node.js, Python, FastAPI, Redis, GraphQL, REST APIs, Playwright.

FinTech / Payment Validation
Payment flow validation, API behavior research, migration rules, E2E evidence, sandbox behavior analysis.


Background

  • PayPal β€” Full-Stack Engineer (2024 – Present)
    FinTech systems, payment migration validation, multi-stack test environments, AI-assisted engineering workflows, identity verification
  • Trip.com (Ctrip) β€” Senior Frontend Engineer (2017 – 2024)
    International content platforms, Node.js SSR, SEO, multilingual content, frontend engineering evolution
  • Earlier β€” Software Engineer β†’ Technical Manager (2005 – 2017)
    Enterprise systems, mobile apps, embedded web, cross-team delivery

Links

Pinned Loading

  1. agentic-souls agentic-souls Public

    AI workflow system with Planner/Evaluator/Specialist trio β€” document-driven, role-based software development

  2. langchain-llm-toolkit langchain-llm-toolkit Public

    LLM toolkit with RAG & multi-model support β€” LangChain-based text generation, chat, and document Q&A

    Python

  3. autogen-asset-analyst autogen-asset-analyst Public

    Multi-agent investment research roundtable built on AutoGen. Four agents (value investor, technical analyst, risk controller with veto power, investment director) debate and reach consensus on port…

    Python 1

  4. nsbp-shop-e2e nsbp-shop-e2e Public

    nsbp shop e2e playwright

    TypeScript 5

  5. autogen-pse autogen-pse Public

    Planner-Specialist-Evaluator three-role agent framework built on AutoGen. Closed-loop delivery: plan β†’ execute β†’ independently verify. Integrates with asset-lens for automated portfolio analysis.

    Python

  6. nsgm nsgm Public

    A powerful full-stack development framework with intelligent code generation

    JavaScript 1