I am Long Nguyen — the engineer and AI consultant behind Rubber Duck Tech Solutions. Funded startups and growing companies hire me as a fractional AI lead: one senior person who owns the strategy, picks the tools that matter, and handles the hands-on integration. AI keeps up with you instead of the other way around.
A fractional Chief AI Officer (fractional CAIO) is a part-time senior AI lead embedded in your company — owning strategy, tool selection, and hands-on implementation without a full-time hire. That's the RDTS model.
Before RDTS, I spent years as the in-house engineer inside two marketing agencies. That is where I learned what teams need from software — and how badly most consultants miss it. I have been writing production software since 2018, pre-LLM, and I have been building with AI models since they were too slow and too expensive for most companies to care.
I run RDTS — Rubber Duck Tech Solutions, the trade name of Long Win LLC, registered in Missouri — as a solo practice. No bench, no offshore subcontractors, no PM between you and the person doing the work. The AI consultant on the call is the engineer wiring the integration.
Proof points from recent engagements: an end-to-end AI music video pipeline I built entirely solo — lyric segmentation, keyframe generation, Seedance video models, multi-model orchestration. Lee De Card — a creator booking platform with AI-assisted matching. Dragon Wagons — a content automation backend. These are not proofs-of-concept. They are production software that runs every day.
My current default kit for artificial intelligence consulting services: Anthropic Claude (frontier for reasoning and long-context work, smaller variants for high-volume classification), OpenAI's flagship GPT where it specifically wins, MCP servers and the Claude Agent SDK for tool-using agents, and custom skills that encode team standards. Harness depends on the team: Claude Code, Cursor, Codex, OpenClaw, Cline, Aider — whichever fits.
If you want more background, the about page goes deeper. If you want to skip ahead, talk to the duck.
Before anything gets built, I audit your current stack and map where AI actually earns its keep. Most teams have five ideas and bandwidth for one. I help you pick the one — or the three, in the right order — and build a roadmap that survives contact with your engineering capacity and your actual business goals. No generic playbooks copied from a conference slide.
The market has hundreds of AI tools, wrappers, and platforms, most of which will not exist in two years. I pick the two or three worth integrating — and name the ten you should ignore. Model selection (which LLM, which tier, which fine-tune if any), harness selection (Claude Code, Cursor, or something custom), and vendor decisions get made on what ships reliably in production, not what looks good in a demo.
Strategy without implementation is a slide deck. I wire the models into your existing stack — your CRM, your CMS, your internal admin, your content database. Custom skills that encode your team standards. MCP servers that give agents first-class access to your tools. Agents that actually run in production, not just in a Colab notebook. See also: AI automation and AI consulting for specific service shapes.
Shipping the integration is half the job. I train your team on what shipped — how to use it, how to prompt it well, what to watch for when it goes wrong, and how to iterate on it without me in the room. Walk-through recordings, short guides, and an async Q&A window are included in every engagement, not billed as consulting hours. If your team is still getting started with AI tools like ChatGPT, see also: ChatGPT for business.
AI moves fast enough that the right answer in January is often wrong by April. On monthly retainer I keep up with the field so you do not have to — new model releases, new tooling, new integration patterns. I flag what changes, update what matters, and keep your stack from going stale. One senior AI person who knows your codebase, at a fraction of the cost of a full-time hire.
When the consulting engagement surfaces a gap that a third-party tool will not fill, I build the app. Internal dashboards, AI pipelines, creator platforms, content automation backends — full-stack production software where AI is a real feature, not the marketing. Source code yours on ship day.
Most engagements combine several of these. A typical first project: strategy + roadmap plus one or two integrations built and shipped. A typical retainer: ongoing model upgrades, new skill development, and quarterly roadmap refreshes. The duck reads which shape fits your situation on the scoping call.
Honest answer: for the right situation, you should. If you need a team of ten, a six-month enterprise rollout, and a PM layer between you and the engineers — hire one of the large AI consulting companies. That is not what I do, and you should not pay solo-consultant prices for it.
Where the large AI consulting firms break down for startups and growing companies: the strategy layer and the implementation layer are different people. A partner writes the roadmap. A junior team implements it. By the time implementation starts, the strategy is already stale — and no one on the implementation team has the context to push back on it. You end up with a perfectly executed plan for the problem you had six months ago.
I am the same person writing the roadmap and shipping the code. When the model behavior surprises us mid-integration, I update the strategy in the same afternoon, not in a follow-up engagement. When a new model drops and changes what was the right call two months ago, I catch it and bring it to you. That is what a fractional AI lead actually is — someone who holds the whole picture, not just the deliverable they were scoped for.
I will not pretend I am the right hire for every job. If your engagement is genuinely 6 hours of scripting, post on Upwork. If you need an army, hire an agency. If you need one focused AI consultant who does the strategy and the build — that is the lane RDTS exists for.
Five-minute scoping conversation with the duck — a custom AI agent I built. You describe what you are trying to accomplish, who is on the team, and what you have already tried. The duck reads the shape of the engagement, names a price in plain English, and sends a brief to both of us. No qualification grid, no funnel.
I send back a one-page scope inside 48 hours: what the engagement covers, what it does not, the format (project, roadmap, retainer, or some mix), the timeline, the deliverables, and the price. Fixed. We align on this before anything starts.
Roadmap first, then integration. I audit the stack, pick the tools, and — in the same engagement — build what needs building. No strategy-only engagements that leave you with a doc and no one to implement it. You see commits as they land and get async check-ins throughout.
On ship day you get the deploy, the source code, the configs, a README another engineer can pick up cold, and a walkthrough recording. A short bug-fix window is included. After that: bring it in-house, keep me on monthly retainer, or call me in for the next round.
Five-minute conversation with the duck — no scoping form, no slide deck. You describe the situation, the duck reads the shape of the AI consulting engagement, names a price, and a brief lands in both our inboxes. If I am the wrong fit, the duck says so. You can also see the rest of the work on the work page.