6 min read
This is an illustrative composite story. “Elena” is not a single real person; her situation and workflow are built from common patterns among independent translators, so you can see how the pieces fit together.
It is 11:40 at night, and Elena is on page 31 of a 60-page product manual, translating from English into Italian. Her coffee went cold two hours ago. The deadline is Thursday. Two more inquiries sit unanswered in her inbox because she simply has no room for them. She has been a freelance translator for twelve years, and for the first time she is wondering whether the job still makes sense.
Six months later, Elena is turning work around faster, quoting with more confidence and, most surprisingly, enjoying the craft again. Here is what changed, step by step, and what any solo service business can borrow from it.
The problem was never the translating
When Elena finally tracked her hours for two weeks, the numbers surprised her. Actual translation, the part she was trained for and loves, took a little over half her working time. The rest disappeared into:
- Writing quotes and answering “how much would it cost?” emails
- Building and checking terminology lists for each client
- First-draft grunt work on repetitive, low-creativity text
- Final proofreading passes on long documents
- Invoicing and chasing late payments
Like many independent professionals, she had been treating AI as a threat to her core skill. The breakthrough came when she started treating it as help with everything around her core skill.
Step one: AI drafts, the professional decides
Elena’s biggest change was adopting a structured post-editing workflow for suitable projects. Post-editing means a machine produces a first draft and a human translator reviews and corrects it to professional quality. It is not a shortcut invented last year; it has its own international standard, ISO 18587, which sets out requirements for the process and the people doing it.
She set herself clear rules about which jobs qualified:
- Good fit: technical manuals, product specs, internal documentation, repetitive e-commerce descriptions.
- Bad fit: marketing slogans, literary work, legal contracts, anything where tone or liability is everything.
For the good-fit jobs, she runs a first pass through a machine translation tool such as DeepL Pro, then edits line by line. On a long technical manual, that shift alone can turn days of typing into days of reviewing, which is faster and, she says, far less draining.
The crucial part: she tells clients. Her quotes now explain which projects use AI-assisted drafting and which are fully human, with different rates. Clients appreciate the honesty, a point we argued at length in why telling clients you use AI is the safer choice.
Step two: a terminology assistant for every client
Every client has favorite words. One insists on “dashboard,” another on “control panel.” Keeping track used to mean digging through old files.
Now, at the start of each new client relationship, Elena pastes a few of their published pages into a general AI assistant and asks: “List the recurring product terms and brand phrases in this text, with the preferred Italian equivalent where one appears. Flag any term used inconsistently.” She reviews the list, corrects it, and saves it as that client’s glossary.
The result is fewer embarrassing inconsistencies and a sharper, more professional deliverable. Before using this approach, she removed client names and anything confidential from the text she pasted, following the kind of habits covered in 5 things solo owners get wrong about putting business data into AI.
Step three: quotes in ten minutes, not an evening
Quoting used to be Elena’s least favorite chore. She would open the file, estimate word counts, guess the difficulty and write a long email. Now she uses a saved prompt that takes the word count, subject area, deadline and her rate card, and drafts a clear quote email with options: AI-assisted draft plus full human edit, or fully human translation.
She still sets the price herself. The AI just handles the writing. That change alone let her answer inquiries the same day, and fast replies win work. Those two unanswered emails from the opening scene? Under the new system, both would have had a quote within the hour.
Step four: a final quality pass she actually trusts
On long documents, tired eyes miss things. Elena added one more step at the end: she asks an AI assistant to compare her final Italian text against the English source and list possible omissions, numbers that do not match, and inconsistent terms. It is a checklist, not a verdict. She decides every change herself.
This step catches the classic late-night errors, a missing sentence, a “15” that became “51,” without replacing her judgment.
What the six months looked like
Because this is a composite story, we are not going to invent precise revenue figures. But the pattern is consistent with what many independent professionals describe when they move from “AI does my job” thinking to “AI handles the edges of my job” thinking:
- More projects accepted, because drafting and admin take less time.
- Clearer pricing, with separate tiers for AI-assisted and fully human work.
- Fewer late nights, which, for a business of one, is not a soft benefit. It is the business.
- More time for premium work, the creative and sensitive projects where a skilled human is irreplaceable and clients pay accordingly.
Elena also joined discussions in her professional community, such as those run by the American Translators Association, to compare notes with colleagues. Learning from peers kept her from both overusing and underusing the tools.
Lessons any solo service business can borrow
You do not have to be a translator for this to apply. Swap “translation” for copywriting, bookkeeping, design or consulting and the playbook holds:
- Track your hours first. You cannot fix time you have not measured.
- Protect your core craft. Use AI on the edges before you use it on the heart of the work.
- Set fit and no-fit rules. Decide in advance which jobs AI may touch.
- Price transparently. Offer an AI-assisted tier and a fully human tier, and explain the difference.
- Keep the human signature. Every final deliverable passes through your judgment.
For another example of the same pattern in a very different field, see how an illustrative solo bookkeeper cut her month-end close in half.
Which part of your work sits on the “edges,” and which part is the heart you would never hand over? Tell us in the comments. The answer might be your best starting point.
