Simon (西门军师)

I studied chemistry — but I’ve been writing code for over twenty years.

My undergraduate and graduate degrees were both in applied chemistry, with a master’s focus on “computer simulation” — essentially running experiments in code rather than in a lab. I completed a second major in computer science, and in the end realized I’d rather stare at a screen and break down a problem into logic than stand at a fume hood waiting for a reaction to finish. So after graduating, I walked straight into software engineering — and never stopped.

The Road I’ve Traveled

The early years were spent at small companies in Shanghai: a residential rental system, an e-commerce backend, a logistics platform integrating data with a US headquarters. The biggest thing I got from those years wasn’t any specific technology — it was the habit of shipping something together with people on the other side of the planet. That habit followed me through almost every job since.

In my mid-twenties, I grew into a Software Development Manager at Entriq, a content SaaS platform company, leading a team of 20–30 people and reporting directly to the VP of Engineering. By the standards of the time, it was a solid position.

Then I made a choice that surprised people: I joined EMC as a Senior Software Engineer — stepping back from managing dozens of people to being an individual contributor. I knew exactly why: I wanted a deeper technical foundation, and I wanted to understand how complex systems inside large companies really grow from nothing. That decision put me through the cycle again — engineer → tech lead → manager → senior manager. Over that decade, my team and I built a suite of enterprise search products capable of returning results across billions of documents in 1–2 seconds. Along the way, I picked up a handful of recognition awards from the CCOE and global headquarters.

After a decade at EMC, I moved to Splunk, a big data company, and shifted dimensions entirely — from building products to building the infrastructure engineers needed: CI/CD platforms, internal productivity tooling, and a localization push that brought WeCom alerting and Chinese word segmentation into the platform.

In June 2020 — the first year of the pandemic, not exactly a “safe-looking” moment — I made the biggest bet of my career: I co-founded a startup. My team and I built a data analytics platform from scratch: from writing the core framework in Python and JS, to eventually leading the integration of LLMs, knowledge bases, NL2SQL, and multi-agent workflows when those things were still very new. For six years, I was probably one of the few CPOs who personally got their hands into both architecture and code.

That chapter has reached a natural endpoint. I’m now starting a new one — founding a startup, applying two-plus decades of accumulated knowledge in technology, product, and team leadership directly to the work I want to do next.

What Has Stayed Constant

Looking back across those twenty-odd years — roles, industries, and tech stacks changed many times — but a few things never really changed.

I’ve always been willing to get in the weeds myself. Whether I was leading a team of dozens or re-starting as an individual contributor, I’ve always believed that you need to have touched the code and the architecture before you have the right to direct others on how to do it.

I’ve always been comfortable collaborating with people who are different from me. From the early years of frequent back-and-forth with US headquarters, to later managing onshore and offshore teams, cross-cultural and cross-background collaboration has never felt like a barrier — it’s just been normal.

And I’ve never been afraid of starting over. From chemistry to computer science, from manager back to engineer, from senior manager at a multinational to co-founder of a startup, to now founding a startup of my own — every “reset,” in hindsight, was worth it.

What I’m Working on Now

Right now, my energy is focused on four things:

Real-world LLM deployment in enterprise contexts — knowledge base integration, NL2SQL, multi-agent orchestration. These aren’t concepts I’ve just read about; they’re things I’ve personally built infrastructure for and written code to make work.

Business process optimization and engineering capability — mapping and automating workflows, acting as an outside advisor to help companies reshape how they work, and applying the methods and hard lessons from complex large-company systems to help smaller, resource-constrained organizations spend less, stumble less, and run more smoothly.

Technology and product judgment — especially the situations that require both technical depth and business intuition simultaneously. That combination is exactly what twenty-plus years of accumulated experience is for.

Startup mentorship and companionship — as a repeat founder, every step from zero to one to commercialization was paid for in real mistakes and real wins. I’m happy to share those experiences with others on the same road. I don’t always have the right answer, but I can at least help you avoid some of the detours.

Let’s Talk

If you’re thinking about how to get LLMs to actually work inside your business, or your company’s workflows and engineering capabilities have room to grow, or you need someone who can sit at the intersection of technical detail and business judgment — or you’re on the hard road of building a startup and want to talk with someone who’s been there — don’t hesitate. Reach out.