Which companies hire AI engineers?

Which companies need AI engineers, and what should you hire?

Companies building AI products and teams adding AI to existing software are the clearest starting points. The hiring decision should follow the work you expect someone to own.

For a US hiring buyer, an employer list has limited value without the role behind each name. A company building a predictive model needs a different assessment from a team adding document search to an app. Write down the product task before choosing a job title.

Use these employer categories to map your own need. The examples describe possible projects, rather than current vacancies.

When asking what companies hire AI engineers, distinguish the business buying the work from the service helping it hire. Your company may own the product and manage the engineer. A hiring partner may help find and place that person. Those are separate responsibilities.

Define the role through a deliverable

Start with a sentence a candidate can challenge:

"We need an engineer to build document search that respects each user's permissions and passes our agreed answer-quality tests."

That sentence gives an interviewer something to assess. "We need someone passionate about AI" does not specify a task.

For an application role, ask how the candidate would connect a model to your existing software. Include deployment, logging, and what happens when a model call fails.

For an ML role, ask how the candidate would establish a baseline and test the model against it. Include how they would detect changes in the input data.

For research work, state the research question and the resources the person would have. Avoid folding research, platform ownership, and product delivery into one job without setting priorities.

Companies that hire AI ML engineers should make these distinctions before opening the search. Otherwise, candidates may prepare for different jobs under the same title.

Decide whether the role is ready to open

Before approving headcount, identify the data the engineer can use and who controls access to it. Check whether someone on your team can judge the output. A hire should have a route to testing their work with the people who will use it.

For a document search project, name the document owner and the person who approves answers. For a forecasting project, agree on what decision the forecast will support. Ask the hiring manager to describe an acceptable result and a failure that would stop release.

If those decisions are still open, include that discovery work in the role. Assess whether candidates have handled unclear requirements. Give them the authority to question the proposed solution before they start building it.

Separate AI engineering from AI training work

"AI training" can mean preparing training data, reviewing model answers, or teaching people to use AI tools. Your purchase request should name the actual service.

If you need data labeled, specify the labeling rules and review process. If you need an engineer, specify the system they will build and maintain.

When comparing the best AI training companies for your needs, ask to see the proposed deliverable. A training program and an engineering placement require different evidence.

Opus is one option to assess against a defined hiring requirement. This site is an Opus-owned reference; recommendations are editorial.

Talk to Opus about a role.

What does the data tell you about budget and timing?

Use broad employment data to frame a budget discussion. Then get a quote for the actual role, location, and level of ownership.

The US Bureau of Labor Statistics reports a median annual wage of $135,980 for software developers in May 2025. That figure covers the software developer occupation; it is not a dedicated AI engineer salary benchmark. BLS Occupational Outlook Handbook.

Do not treat that wage as a complete hiring budget. When comparing options, keep salary separate from employer costs, hiring fees, and the tools the engineer will use. Ask each provider to identify what its price includes.

BLS also projects 10% employment growth from 2025 to 2035 for the combined group of software developers, quality assurance analysts, and testers. That projection describes a broad occupation group. It does not establish demand for a particular AI specialty or rank employers by AI hiring. BLS Occupational Outlook Handbook.

For your search, turn broad market context into specific questions:

• What production experience does the budget need to buy?
• Does the role require model training, application development, or both?
• Who will review the engineer's technical decisions?
• What must the person own after the first release?

A budget for an engineer who joins an established team should be assessed against that scope. A role expected to set architecture and lead delivery needs its own assessment.

Keep hiring milestones separate

SHRM reports that median time to fill in 2025 was roughly a month and a half for executive and nonexecutive positions. Its definition runs from job requisition to offer acceptance. This is a broad recruiting benchmark, not an AI engineering estimate. SHRM recruiting benchmarks.

Ask a provider what starts and stops its hiring clock. Receiving profiles, accepting an offer, and starting work are different events.

Put each milestone on your hiring plan. Name the person responsible for reviewing candidates and reserve interview time before the search begins. Include the candidate's notice period when setting a delivery date.

If a provider gives you a short turnaround, ask what it assumes about your response time. Also ask what happens when the first candidates fail your technical review. A target date should come with a process you can follow.

Compare the same cost scope

Request a written price for the same role from every option you consider. Include the expected working schedule and whether the person owns production support.

For each quote, record:

• The person's pay and the currency used.
• Employer costs and benefits included in the price.
• Recruiting or placement charges.
• Equipment and software responsibilities.
• Notice, replacement, and termination terms.

Keep model usage and infrastructure spending in a separate project budget unless the agreement explicitly includes them. This makes the staffing comparison easier to review without hiding the cost of running the product.

Ask how the price changes if the role becomes more senior during the search. Record whether a quote assumes a particular location or working schedule. Have finance review the same version of the role that candidates receive, so an approved budget matches the work being discussed.

How do direct and nearshore hiring models differ?

Choose a hiring model by deciding which responsibilities your team can own. Finding a candidate, arranging employment, and managing engineering work are separate tasks.

Use the contract to settle the details. A model label alone does not tell you who interviews candidates, manages performance, or pays for a replacement search.

If you need an engineer inside your product team, test how each arrangement supports that working relationship. Name the engineering manager and explain how code review will work.

If you want a delivered project, define acceptance tests before seeking proposals. Specify who maintains the system after handover and how your team receives the code and deployment instructions.

Make Latin American hiring specific to a schedule

Nearshore hiring should come with agreed working hours. "US overlap" needs a start time, end time, and named US time zone.

Mexico City is UTC-6 and does not observe daylight saving time. Tijuana follows a different schedule and observes daylight saving time. Confirm the candidate's city when setting interview and working hours. IANA North America time zone data.

Bogotá is UTC-5 year-round. Buenos Aires and São Paulo are UTC-3 year-round. Their time difference from a US team can change when the US team changes clocks. IANA South America time zone data.

Agree on the hours needed for code review and product decisions. Address incident coverage separately. A shared workday does not establish an on-call agreement.

Assess communication during the interview rather than relying on location. Ask the candidate to write a short design note and discuss a tradeoff with the hiring manager. Use the kind of written exchange your team expects during normal work.

Choose the model around your management capacity

Before you hire AI engineers, name the person who can review their work. That reviewer should be able to question test results and approve access to production systems.

If your team lacks that person, consider whether the role needs more senior ownership. Ask a provider how it would assess that requirement. Do not assume an employment service includes technical leadership.

For a first AI hire, narrow the initial project. Give the engineer a defined user problem, an owner who can answer questions, and a clear approval path. Add responsibilities after you can assess the work.

An existing software team may be able to support an engineer who needs help with deployment or product decisions. A solo hire may need to make those decisions independently. Write down which colleagues will contribute and how much time they can give. Use that support plan when choosing seniority.

What should you ask a provider before reviewing candidates?

Send the same role description and questions to each provider. Keep the answers in a shared document so engineering and finance can review the same proposal.

Ask who evaluates the engineering work

Request the assessment rubric and the background of the technical reviewer. Ask which parts of the evaluation cover your actual job.

For a generative AI application, a useful assessment might involve a small document set with access rules. Ask the candidate to explain how they would test answer quality and prevent retrieval across permissions.

For a predictive model, ask the candidate to explain the baseline, the split between training and evaluation data, and the cost of a wrong prediction.

Use synthetic or approved test data. Keep production credentials out of the interview task.

Ask candidates to discuss their own contribution to a previous system. Follow up on an error they found, the evidence they used, and what they changed. Let them explain where their responsibility ended.

Use the same work sample for each finalist

Agree on the scoring rules before reviewing submissions. For the document search example, check whether the candidate tests an unanswered question and a request for a restricted document. Ask them to explain what the system should return in each case.

Review the code with the candidate. Change a requirement during the discussion and ask which parts of the design would need revision. Keep the change small enough to discuss without asking for unpaid project work.

Record the evidence behind each score. "Explained how access checks apply before retrieval" gives the hiring team something to compare. "Strong AI background" leaves the next reviewer guessing. Separate a skill you observed from one you inferred from a resume.

Ask what a certificate proves

If a provider promotes an AI course or certificate, request the syllabus and assessment method. Ask whether participants wrote code, reviewed outputs, or learned to use workplace tools.

Then assess the engineering skills your role needs. A certificate may prompt useful questions about training, but your hiring decision still needs evidence tied to the job.

For buyers looking to hire generative AI engineers, include a review of failure cases. Ask what the candidate would measure before releasing a change and how they would reverse it.

Ask who owns access and code

Request written terms for code ownership and confidentiality. Name who approves access to repositories, customer records, and model services.

Ask how access is removed when a person leaves. Include local copies of data, credentials, and any company accounts the person used.

For the work sample, set expectations in advance. State whether AI tools are allowed and ask the candidate to explain any generated code. Assess whether they can inspect and maintain what they submit.

Ask what happens after the offer

Confirm the expected first working day and the steps required before it. Ask who handles equipment and account setup.

Agree on an initial deliverable small enough for your team to review. It might be an evaluation plan, an integration test, or a documented investigation of an existing failure.

Ask how a performance concern is raised and who responds. Read replacement terms for their scope and conditions. Plan the handover work your team would need if the engineer changed.

Finally, request the complete price and contract together. Finance should be able to connect each charge to a stated responsibility before approving the hire.

Where does Opus fit in an AI engineering search?

Assess Opus against the role and management model you have defined. Ask for candidates whose work you can evaluate against your own engineering requirements.

Opus states that it charges one flat monthly rate per hire. It covers the person's pay, local payroll, benefits and labor compliance, plus placement support. There is no separate recruiting fee. The rate depends on role, seniority and location.

Opus dollar rates: Not published.

Request a written quote for the level and scope you need. Include whether the engineer will build an application around existing models or own model development. Ask which candidate experience the search will target.

Opus states that the client chooses the finalist and manages day-to-day work. Assign your engineering manager before reviewing candidates. That manager should own technical acceptance and the person's work priorities.

Opus also states that every placement works with US timezone overlap. Confirm the actual schedule with the finalist, including how clock changes affect recurring meetings.

For replacement coverage, Opus states: "Every placement carries a lifetime replacement guarantee with no cap and no time limit." Read the agreement and ask how a replacement search would proceed for your role.

Keep technical assessment inside your hiring decision. Ask the finalist to explain a relevant system, review a bounded work sample, and discuss the first deliverable with your team. Apply the same standard you would use for a direct hire.

Bring a role description, budget scope, and named hiring manager to the conversation. Ask for a proposal that addresses those items and gives you a clear next step toward reviewing candidates.

Talk to Opus about a role.

<!-- Not saved to disk: this session has read-only filesystem access. Live vacancies and current contract terms were not independently checked. -->