Work on AI that actually works
Bombyx Labs is hiring practical AI builders.
Not prompt magicians.
Not people who want to sell big strategy decks and disappear.
We are looking for people who can turn real business knowledge into working AI systems: website assistants, lead follow-up helpers, internal knowledge tools, customer response workflows, question reporting, ecommerce product-finding helpers, and safer first-pass systems that keep people in control.
The assistant is what people see.
The business brain is what makes it useful.
That is the work.
What we build
Most businesses do not need another blank AI tool.
They need AI that understands the business well enough to help with real work.
That means the system needs to know the company’s services, policies, offers, customer questions, tone, handoff rules, edge cases, and owner judgment.
At Bombyx Labs, we build that knowledge layer first. Then we use it to power practical AI helpers and custom systems.
We call that building from the brain.
Who we are hiring now
We are hiring for a small set of roles that support practical AI implementation.
Some roles may start as contract, part-time, or project-based. Some may grow into deeper long-term work. What matters most is fit, judgment, and the ability to help real businesses get useful AI into actual use.
Forward-Deployed AI Engineer
This is the hands-on builder role.
A forward-deployed AI engineer works close to the client, learns how the business actually runs, builds the missing system, connects it to the right tools, and improves it from real use.
You might be a fit if you can write production code, work with application programming interfaces, understand messy workflows, ask good questions, and explain technical tradeoffs in plain English.
What you might build
- Website sales assistants
- Lead follow-up systems
- Internal knowledge helpers
- Customer response workflows
- AI-assisted reporting tools
- Custom integrations with websites, forms, email, customer relationship management tools, or ecommerce systems
- Safe handoff paths between AI and humans
You are probably a fit if
You like working directly with real business problems.
You can move between code, workflow, and customer conversation.
You can build the first useful version without turning everything into a six-month architecture project.
You know when AI should act, when it should ask, and when it should stop.
Plain version: you turn one real business bottleneck into one useful working system.
Business Brain Architect
This is the knowledge role.
A useful AI helper needs more than a prompt. It needs the business brain: services, policies, offers, frequently asked questions, tone, handoff rules, examples, edge cases, and the judgment that usually lives in the owner’s head.
A Business Brain Architect helps gather, organize, clean, and shape that knowledge so AI can use it safely.
You might come from writing, operations, customer support, user experience, documentation, training, research, content strategy, or knowledge management.
What you might work on
- Turning scattered company knowledge into clear source material
- Organizing services, policies, offers, and customer questions
- Writing answer rules and review notes
- Capturing owner judgment and edge cases
- Finding gaps in the business brain
- Improving assistant answers from real conversations
- Helping the AI know when to answer and when to hand off
You are probably a fit if
You can turn messy information into clear structure.
You notice when an answer is technically true but not actually helpful.
You care about clarity, source quality, and real-world judgment.
You understand that AI is only useful when the business knowledge underneath it is strong.
Plain version: you help teach the system what the business already knows.
AI Workflow Builder
This is the automation role.
Many businesses do not need a brand-new app. They need one repeated workflow to stop eating the week.
An AI Workflow Builder connects tools, builds automations, drafts safe first-pass systems, and helps move work from “someone has to remember this” to “the system helps.”
You might work with tools like Zapier, Make, n8n, Airtable, Google Workspace, forms, customer relationship management systems, scripts, application programming interfaces, and lightweight apps.
What you might build
- Lead routing
- Quote intake
- Customer response drafts
- Weekly owner reports
- Question reporting
- Internal search helpers
- First-pass summaries
- Review-and-approve workflows
- Simple automations between existing tools
You are probably a fit if
You like making messy work smoother.
You can spot repeated tasks hiding inside a business process.
You can build lightweight systems without making them fragile.
You care more about whether the workflow works than whether the tool is trendy.
Plain version: you help AI take the first pass while people stay in control.
AI QA & Agent Operations Reviewer
This is the trust role.
AI systems need testing. Not “does it sound good?” testing. Real testing.
Does the assistant answer from the right source? Does it follow the business rules? Does it make things up? Does it ask for the right details? Does it hand off when the situation is sensitive? Does it protect the business from a confident wrong answer?
That is the work.
What you might test
- Website assistant answers
- Lead capture flows
- Human handoffs
- Knowledge gaps
- Unsafe or unsupported claims
- Tone and clarity
- Repeated failure patterns
- Real customer questions
- Review loops and correction workflows
You are probably a fit if
You are detail-oriented without being slow.
You are skeptical in a useful way.
You can find edge cases before customers do.
You care about making AI helpful, safe, and boringly reliable.
Plain version: you help make AI useful without handing the keys to a black box.
AI Implementation Partner
This is the adoption role.
A good system still needs people to understand it, trust it, and use it.
An AI Implementation Partner helps clients get from “interesting idea” to “this is part of how we work now.” That may include onboarding, workshops, team training, setup calls, documentation, feedback reviews, and helping clients choose the next useful thing to build.
What you might help with
- Client onboarding
- Business brain setup
- AI workshops
- Team training
- Use-case discovery
- Follow-up reviews
- Documentation
- Feedback collection
- Helping clients improve their assistant over time
You are probably a fit if
You can teach without lecturing.
You can simplify without dumbing things down.
You are comfortable with business owners who are interested in AI but overwhelmed by the noise.
You can help people see the next practical step.
Plain version: you help people put the system to work.
[Apply for this role]
Product & Conversation Designer
This is the clarity role.
AI helpers need more than answers. They need a good path.
What should the assistant ask first? When should it suggest a next step? What should the button say? When does the user need reassurance? What does the business need to know before following up?
A Product & Conversation Designer shapes the experience so the AI helper feels clear, useful, and human.
What you might shape
- Website assistant flows
- Lead capture questions
- Product-finding paths
- Customer handoff language
- Assistant tone
- Review screens
- Owner reports
- Simple interfaces for business teams
You are probably a fit if
You think carefully about how people move through a decision.
You care about plain language.
You can make a buying path feel less confusing.
You understand that good design is not decoration. It is how people know what to do next.
Plain version: you make the AI helper easier to trust and easier to use.
Ecommerce AI Systems Builder
This is the product-finding role.
Some stores lose sales because customers cannot figure out what to buy.
The catalog is too complex. The compatibility rules are too specific. The product pages are too long. The customer does not want to make the wrong choice.
An Ecommerce AI Systems Builder helps turn product confusion into buying confidence.
What you might build
- Product-finding helpers
- Compatibility checkers
- Shopify workflows
- Catalog cleanup systems
- Buying guidance assistants
- Product question-and-answer tools
- Customer confidence flows
- AI helpers for complex product lines
You are probably a fit if
You understand ecommerce.
You care about the buying path.
You can work with product data, filters, variants, compatibility rules, or customer questions.
You know that “Will this work for me?” is not just a support question. It is often the sale.
Plain version: you help customers choose the right thing without needing a support ticket.
How we work
We start with one real business problem.
Not a giant artificial intelligence roadmap.
Not a vague transformation plan.
One workflow. One bottleneck. One useful first version.
Then we build from the brain: the business knowledge, rules, tone, examples, and judgment that make the helper useful.
We test it.
We improve it.
We keep people in control.
Then, if it works, we build the next piece.
What we care about
We care about useful systems.
We care about clear writing.
We care about customers getting better answers.
We care about owners seeing what is happening in their business.
We care about lead follow-up that does not die in an inbox.
We care about AI that knows when to stop.
We care about small businesses, community organizations, consultants, ecommerce stores, and professional teams that need practical help, not hype.
You might be a good fit if
You like practical work.
You can talk to business owners without hiding behind jargon.
You care about outcomes, not just tools.
You can work with messy information.
You can explain what you built.
You can spot where AI should stop.
You would rather build one useful system than make a flashy demo.
You believe generic is where value goes to die.
You are probably not a fit if
You want to hand clients a strategy deck and disappear.
You think AI should replace every human judgment call.
You use buzzwords before you understand the workflow.
You need every problem to be perfectly defined before you start.
You care more about the tool than the business result.
You cannot explain your work in plain English.
How to apply
Send us a short note.
Tell us which role sounds closest, what kind of work you have done, and one real workflow you would be excited to improve with AI.
Helpful links are better than a polished resume.
Show us what you have built, fixed, organized, taught, designed, tested, or made easier.
Please include
- Your name
- Your email
- Where you are based
- The role you are applying for
- Whether you are looking for full-time, part-time, contract, or project-based work
- Links to work, writing, code, portfolio, demos, case studies, or examples
- One workflow you have improved, automated, clarified, designed, or tested
- Why Bombyx Labs sounds interesting to you
[Apply now]
For business owners reading this
This is the kind of team we are building for you.
Not a chatbot vendor.
Not a pile of disconnected tools.
A small, practical build team that can learn your business, build from your knowledge, and help AI do useful work without taking people out of the loop.