My name is Chris Lam. I'm an AI applications engineer residing and working in Bentonville, AR. I design and ship multi-model, multimodal AI pipelines, internal platforms and automations from first prototype to production, at enterprise scale. Because I want teams to get home to their families, not their inboxes.
I map how the work actually happens — who touches it and where it stalls — before choosing a model or a tool.
A working pipeline in week one beats a perfect design in month three. Real usage decides what gets built next.
Validation, health checks and plain logging, so the system keeps running after the demo and the people using it stop noticing it.
Each one started with a process that cost people hours, and ended in production.
Phone photos of an illustration-only assembly manual in; sequenced written instructions out. One model identifies the parts, another model interprets each step, a validation loop checks the order.
For a paid-search agency: scraped 2,000+ dynamically rendered landing pages, pulled structured data with a language model and generated assets for 2,000+ unique Google Search ads. A daily health check auto-pauses any ad whose page goes down.
Syncs multicam footage and audio, throws out the unusable takes and assembles a first-draft edit — so editors start from a cut instead of a pile of clips. Then, a custom Remotion editor streamlines the final cut.
A Python QA tool that calculates a turntable’s real rotational center in real time using computer vision, then overlays crop boundaries and live centering guides for camera operators.
A first-of-its-kind Airtable architecture that automatically documents, version-logs and tracks the live state of production Interfaces — something the platform doesn’t do natively.
Goal bingo for people who’ve given up on New Year’s resolutions. Build a card each season, invite friends to add squares and call which ones you’ll hit, and attach receipts (photos) when you daub one. Randomized squares mean no ranking your goals — they just stay quietly in play.
The app behind the login on this very domain — my personal operating system. It runs its own MCP server, so Claude can read and write to it with me.
Side projects in various states of done. I’d rather ship a small thing people use than plan a big thing nobody sees.
An AI journaling and self-advocacy app. Write what happened in your own words; it helps you turn that into something clear you can bring to the appointments and conversations that matter.
A Chrome extension + webapp that documents Airtable Interfaces — the layer that Airtable’s API can’t see. For every builder who’s inherited a base with zero docs.
Spreadsheets emailed back and forth? Someone copy-pasting 400 rows every Monday? I collect those. Tell me about yours.
Growth marketing and my own film company first. For the last five years: AI engineering, internal platforms, enterprise Airtable and front-end.
Internal AI tooling for a high-volume product photo and video capture operation — multimodal pipelines across pre-production, production and post. Retained through a restructuring from ~200 people to 25.
Enterprise Airtable systems used across multiple teams — one system for planning, approvals and delivery in place of scattered documents, plus change detection that flags only what matters.
An end-to-end AI content pipeline for a paid-search agency’s ad operation at scale.
Technical lead for product data and publishing across multiple brand sites: a hybrid product-information platform (Airtable → HubDB → HubSpot CMS + React) at a fraction of enterprise PIM cost, plus AI-generated product pages.
Front-end lead on an LLM-powered business-intelligence platform: chat interface, prompt wizard and a branded component library.
Rebuilt an agency’s operations in Airtable, with integrations and scripted workflows that saved about 32 hours a week.
Led acquisition that brought in 80,000+ new customers at target cost, and mentored a performance marketing team.
Founded and ran a video production company, directing branded and commercial work end to end. Before that, freelance web development for small businesses.
Multi-model orchestration · multimodal pipelines · computer vision in Python · structured extraction · prompt engineering · embeddings & vector search · OpenAI & Anthropic (Claude) APIs · MCP · [evals & testing — if true]
TypeScript · JavaScript · Node.js · Next.js · React · Python · REST APIs & webhooks · Chrome extensions · Airtable Extensions SDK · Git
Postgres & SQL · Prisma · Vercel · Clerk auth · Vercel Blob · MCP servers · [observability — if true]
Airtable architecture, scripting & interfaces · Retool · HubSpot CMS & HubDB · web scraping · data sync pipelines
MBA, Leadership & Org Development — Oklahoma Christian · BA Psychology, magna cum laude — Oklahoma Baptist · Certified Airtable Builder
I came to engineering from psychology, film and marketing — so I build for the people using a system, not just the system.
My psychology degree taught me how people actually work. Running a film company taught me how to ship under pressure. Years in marketing and dev taught me that the best tool is the one people forget they’re using.
That range is why my systems get adopted: I start with the people doing the work, get the real requirements on the table, and ship something they can rely on.
First-gen son of an immigrant, raised in a small Oklahoma town, where I spent most of my childhood learning magic tricks.
A psychology degree, then an MBA. Freelance WordPress sites, then my own video production company. Growth marketing for a national e-commerce brand — and my first serious engineering roles.
Moved to Bentonville to work for Sam’s Club, then built AI tooling in retail’s backyard. Open to remote roles and onsite in NW Arkansas.
Outside of work I build with my hands, too.
I’m open to remote roles and onsite in NW Arkansas — AI application engineering, automation and creative technology. The inbox is always open.