Side Hustles for Engineers: Ten States Protect You, Nevada Won't

The standard list of side hustles for engineers starts with advice that can forfeit your intellectual property. Write a software application on nights and weekends, launch a software-as-a-service product, or publish a specialized CAD library. None of those articles examines who owns the asset once you launch it. If you work as an engineer in the United States, most engineering employment agreements include an invention assignment clause. That clause can claim full ownership of your side project before you make your first dollar.

Finding sustainable side income requires sorting opportunities by intellectual property risk. When you build a product, your employer might own it under your contract. When you sell engineering services to the public, state licensing boards require professional engineering credentials and firm registrations that do not apply to your day job. Freelance marketplaces pay rates that fall below your regular salary. The clean alternative is selling judgment where you own nothing, build no standalone product, and complete work delivered through platforms to AI labs. Those evaluation roles can match or beat your salaried hour, without triggering the clauses in your employment contract.

Disclosure: Skillora earns a referral fee from micro1, Mercor, and Alignerr if you sign up through our links and work on their platforms. We earn nothing from DataAnnotation or Outlier. Our referral agreements cost you nothing, and the pay rates cited below come directly from published platform listings.

The bar every engineer side hustle has to clear

Your primary engineering salary establishes a hard financial floor. Taking on side work that pays less than your salaried hourly rate means accepting a voluntary pay cut. According to the May 2025 Occupational Outlook Handbook from the Bureau of Labor Statistics, median earnings across core engineering disciplines break down to these hourly equivalents (calculated as annual salary divided by 2,080 working hours):

• Software developers: $135,980 per year, or about $65 per hour.
• Electrical and electronics engineers: $125,040 per year, or about $60 per hour (electrical engineers median $120,630; electronics engineers except computer median $130,220).
• Mechanical engineers: $104,110 per year, or about $50 per hour.
• Civil engineers: $100,840 per year, or about $48 per hour.

For comparison, the median wage for all US workers in May 2025 was $50,980. If you are an engineer earning $50 to $65 per hour at your desk, spending your free time bidding for $30 per hour work on general gig platforms makes no mathematical sense. Your side hustle must beat your salaried baseline while operating cleanly outside your employment agreement.

Who owns what you make: the question every list skips

Most engineering employment contracts include a Proprietary Information and Inventions Agreement (PIIA). This agreement governs who owns your creations, code, and inventions during your employment tenure. The contract text determines ownership unless a state statute steps in to protect you. This guide provides general information, not legal advice. The specific language in your signed employment agreement governs your relationship, and an employment lawyer can review your contract.

Under California Labor Code 2870, an invention assignment agreement cannot touch an invention developed entirely on your own time without using your employer's equipment, supplies, facilities, or trade secret information. However, California law carves out two critical exceptions where the employer still claims ownership:

1. Inventions that relate, at the time of conception or reduction to practice, to the employer's business or actual or demonstrably anticipated research or development.
2. Inventions that result from any work you performed for your employer.

Under California Labor Code 2872, employers must provide written notice of this statutory limit in any employment agreement executed after January 1, 1980. Any contract clause that attempts to reach further than Section 2870 permits is void as against public policy.

Nine other states have enacted statutes protecting an employee's own-time inventions with similar carve-outs:

• Delaware (19 Del. C. 805): Protects projects completed on personal time without company assets.
• Illinois (765 ILCS 1060/2, Employee Patent Act): The employee bears the burden of proving that an invention qualifies for the statutory protection. The employer must give written notice in agreements signed after January 1, 1984, and cannot require an overly broad, void clause as a condition of employment.
• Kansas (K.S.A. 44-130): Restricts assignment agreements reaching beyond company business.
• Minnesota (Minn. Stat. 181.78): Provides the standard own-time, own-equipment carve-out.
• New Jersey (N.J.S.A. 34:1B-265): Protects personal inventions built without employer resources.
• New York (Labor Law 203-f): Signed into law on September 15, 2023, barring employers from claiming inventions made on personal time with personal resources.
• North Carolina (G.S. 66-57.1): Limits employer claims on off-hours development.
• Utah (Utah Code 34-39-3, Employment Inventions Act): Protects personal inventions that do not meet the statutory definition of an employment invention.
• Washington (RCW 49.44.140): Protects off-the-clock employee work that avoids company resources.

Nevada runs in the opposite direction. Under Nevada Revised Statutes (NRS 600.500), an employer is the sole owner of any patentable invention or trade secret developed in the course and scope of employment that relates directly to the work, unless an express written agreement states otherwise. In states without protective statutes, your signed contract decides ownership entirely.

Even in protected states, software side projects face severe legal vulnerabilities. The statutory protection demands that three conditions hold simultaneously: you must build on your own time, using your own equipment, on something unrelated to your employer's business. The phrase "relates to the employer's business or demonstrably anticipated research" catches software engineers routinely. A software employer's business is usually broad, which is why this exception catches software side projects most often. If your SaaS tool touches any problem your employer evaluates, your employer may claim the asset. Building on a company laptop fails the equipment test, because the statutes require that no employer equipment was used.

Tier one: things you build, and might not own

Tier one side hustles center on creating intellectual property: standalone web applications, SaaS businesses, developer plugins, hardware designs, automation scripts, and CAD model libraries. Every guide tells you these create passive income and equity. None mentions that they carry the highest legal exposure for an employed engineer.

If you build an independent product while employed full-time, take concrete precautions before writing code or drafting schematics:

• Read your agreement carefully. Check your offer letter and your signed PIIA for invention assignment clauses, non-compete covenants, and outside-work restrictions.
• Verify statutory notices. In states like California and Illinois, check whether your employer included the mandatory written disclosure outlining your statutory carve-outs.
• Maintain strict physical separation. Never run side project code on your employer's laptop. Never access project repositories from your corporate network. Buy your own hardware, set up independent email accounts, and pay for personal cloud hosting accounts.
• Stay away from your employer's market. If your employer builds warehouse management systems, do not build an inventory tracking tool. Build entirely outside your employer's sector.
• Request a written carve-out. Submit a formal written disclosure to your employer's legal or HR department detailing your side project and requesting an explicit waiver. If they decline to clear the project in writing, you know the asset is contested before you build it.

Because of these ownership risks, tier one projects rarely offer a simple path to supplementary income.

Tier two: work you sell to someone else

Tier two side hustles involve selling technical work to clients: traditional freelancing, software contracting, direct consulting, expert witness testimony, and technical writing. In this category, your client receives the deliverables, sidestepping the risk of building an asset your employer can seize. However, tier two brings moonlighting clauses, conflict-of-interest policies, and regulatory licensing gates.

For mechanical, civil, and electrical engineers, state licensing laws impose strict barriers. As detailed in Skillora's guide to freelance mechanical engineering, most corporate engineers work under the industrial exemption. This exemption allows engineers to design internal products for their employer without holding a Professional Engineer (PE) license. The industrial exemption attaches to the work, not to your degree. Once you market engineering services directly to third-party clients, you are offering engineering services to the public. According to the National Council of Examiners for Engineering and Surveying (NCEES), only a licensed engineer may prepare, sign, seal and submit engineering drawings.

Commercial rules present another hurdle. Offering engineering services to the public usually requires formal firm authorization. Harbor Compliance reports that 34 states require engineering entities to hold a Certificate of Authorization, while 8 additional states enforce separate firm registration mandates. Setting up an independent corporate entity and securing professional liability coverage introduces substantial overhead.

Traditional freelance platforms rarely justify this compliance burden. On Upwork, the published median rate for mechanical engineers sits at $35 per hour, within a typical range of $30 to $50 per hour. Earning $35 per hour as a freelancer compares poorly against a $50 per hour salaried wage. Break-even works out around $96 an hour once overhead is counted, per the mechanical engineer guide. While independent consulting fee schedules show project engineers holding a PE billing $120 to $155 per hour, and firm principals commanding $250 to $350 per hour, landing those engagements requires extensive client development.

Other tier two options feature specific trade-offs:

• Expert witness consulting: The 2026 ExpertPages survey of over 400 experts found that the average expert witness fee sits at $465 per hour across all disciplines, with rates spanning from under $200 to over $1,500 per hour (referenced in Skillora's analysis of physician side gigs). Conflicts checks and your employer's outside-activity rules apply.
• Technical writing: Writing documentation or technical content offers clean boundaries. According to the BLS, technical writers earned a median annual wage of $90,390 in May 2025. This translates to roughly $43 per hour, which falls below average engineering pay.
• FE and PE exam tutoring: Teaching candidates to pass the Fundamentals of Engineering or Professional Engineering examinations requires no professional license. However, finding individual students takes continuous marketing effort.

Software engineers face no state PE licensing gates. However, client freelancing still triggers employer non-compete terms and outside-activity policies, while exposing you to demanding client schedules.

Tier three: judgment you sell, where nothing is yours to claim

Tier three includes side hustles where you sell professional evaluation and expertise rather than custom products or deliverables. You do not build a proprietary software product, so your employer's invention assignment clause has no asset to claim. You do not provide regulated engineering services to the public, so you need no PE seal or firm registration.

AI training work sits at the center of tier three. When you evaluate an artificial intelligence model, your intellectual work product is assigned to the lab under the platform's contract. You build no product of your own. The work involves reviewing technical reasoning, annotating edge cases, drafting technical problems, and verifying algorithmic logic. It is purely evaluative work.

Other tier three options include paid expert network phone consultations. As covered in our freelance mechanical engineer guide, expert networks pay $100 an hour and up for specialized industrial knowledge. Call volume is somebody else's pipeline you cannot plan a month around.

AI evaluation platforms provide hourly pay that you can schedule around your primary job, though volume is project-based. You complete tasks on your own computer, during your personal hours, without marketing yourself to corporate clients or negotiating contracts.

AI training jobs for engineers: what they pay

AI model developers require technical domain experts to evaluate complex model outputs. For technical professionals, this work involves writing and debugging complex code, creating benchmark problems, auditing AI-generated software architectures, assessing multi-step mathematical and physical reasoning, and grading model outputs against detailed rubrics.

Platform pay data reflects active listings posted by September 6, 2026, across major AI evaluation providers. We differentiate between typical ranges (the median floor to median ceiling of active postings) and top listings (the highest specific rates advertised). Every rate cited in the table below links directly to live postings on our job board.

Examining these rates against BLS baseline salaries reveals immediate financial advantages for hardware and civil disciplines. A mechanical engineer earns a median salaried wage of about $50 per hour, and a civil engineer earns about $48 per hour. For engineering roles outside software, the typical ceiling on micro1 and Mercor is roughly double their salaried hour, reaching $90 to $100 per hour, with top listings reaching $110 to $130 per hour (Alignerr's engineering roles pay less, $35 to $60 per hour). For engineers in physical disciplines, expert-level AI evaluation provides strong hourly earnings without requiring a PE license.

Software engineers straddle a wider compensation band. Against a salaried median of $65 per hour, standard software evaluation tasks on Alignerr start at $50 to $90 per hour. High-complexity software roles on micro1 and Mercor pay substantially more. micro1 lists Senior Software Engineer and Data Scientist roles paying $245 to $280 per hour. Mercor offers Legacy Codebase Migration Expert roles paying $200 per hour, and Offensive Security research roles paying $200 to $250 per hour. Conversely, basic robotics data annotation listings pay typical rates of only $10 to $24 per hour. Engineers should avoid low-tier data labeling and focus strictly on expert-level evaluation roles.

Active listings posted by September 6, 2026, highlight the diversity of specialized technical roles available:

• micro1 technical roles: Senior Software Engineer ($245 to $280/hr), Data Scientist ($245 to $280/hr), AI Software Engineering Domain Expert ($100 to $200/hr), AI/ML Engineer ($70 to $200/hr), Senior Software Engineer ($100 to $150/hr), CAD Expert ($80 to $130/hr), Autodesk Fusion CAM Expert ($80 to $120/hr), CAM Programmer ($80 to $120/hr), SolidWorks Specialist ($40 to $100/hr), and Hardware Expert ($50 to $100/hr).
• Mercor technical roles: Legacy Codebase Migration Expert ($200/hr), Machine Learning Engineer Talent Network ($70 to $250/hr), Frontend, Backend, Full-Stack and DevOps/Platform Engineer Talent Networks ($70 to $150/hr), Mechanical Engineers and Industrial Engineers ($70 to $110/hr), Radio Frequency and Electromagnetic Engineer ($80 to $95/hr), Engineering Simulation Specialist ($60 to $90/hr), Electrical Design Specialist ($70 to $85/hr), and Offensive Security research expert ($200 to $250/hr).
• Alignerr technical roles: Machine Learning Evaluation Specialist ($200 to $400/hr), Robotics ML Expert ($100 to $150/hr), Code Review Specialist and Software Testing Analyst ($40 to $120/hr), along with Mechanical, Electrical, Computer, and Environmental Engineering evaluation roles ($35 to $60/hr).
• Outlier domain postings: Advertised domain ceilings reach up to $150 per hour for Electrical Engineering Experts and ML/AI Experts.
• DataAnnotation technical roles: Software Engineer and Machine Learning Engineer listings range from $75 to $150 per hour.

To view active listings across these specialties, explore our dedicated board hubs for software engineering roles, data and machine learning positions, engineering roles outside software, and robotics listings. You can also review general roles across the broader AI training board or read our overview of the best AI training jobs.

All of these engagements operate on an independent contractor basis. You work remotely on flexible, asynchronous schedules, selecting your hours once matched to a project. Contractors receive gross payments with no taxes withheld, making tracking self-employment obligations essential.

How AI training platforms screen engineers

AI training platforms maintain rigorous automated screening processes to evaluate technical expertise and communication clarity. Understanding how each platform screens applicants ensures you do not waste your application attempts:

• micro1: Applicants complete an interview with Zara, a conversational AI interviewer that asks adaptive technical questions based on your responses. For technical roles, micro1 deploys its Ava proctoring system during coding evaluations. Ava tracks gaze patterns, tab switching, external monitor setups, and browser extensions to generate an integrity score. Reporting indicates that an integrity score below roughly 70% results in rejection, while using AI copilots to inject code triggers an immediate failure. micro1 publishes no retake policy, though they offer a free Interview Prep tool. Learn more in our micro1 AI interview guide.
• Mercor: Screening centers on a 20-minute conversational AI interview. Applicants receive three total attempts shared across all platform applications, with Mercor saving and scoring only your most recent submission. The platform provides an unlimited free practice interview feature so you can refine your delivery. Review our Mercor AI interview guide for detailed preparation strategies.
• Alignerr: Requires an automated AI screening interview followed by domain-specific technical qualification tasks. If you fail an assessment for one role, you can apply to a different domain listing the following day. Alignerr runs two payment models on projects, and only one pays for time.
• DataAnnotation: Features a single unpaid initial assessment with only one attempt permitted. Software candidates must complete a dedicated coding assessment to access technical projects.

To maximize your chances of getting approved across platforms, follow our step-by-step AI training platform screening order. Before taking a live evaluation, test your delivery using Skillora's scored mock interview tool, which uses an AI interviewer to generate dynamic technical follow-ups from your answers.

The confidentiality line

While AI evaluation avoids invention assignment issues, it introduces distinct confidentiality obligations. You must observe strict operational boundaries to protect your primary employment:

• Never share employer assets. Never paste proprietary code, technical schematics, CAD files, customer datasets, or internal strategy documents into an AI evaluation prompt or grading form. Doing so violates your confidentiality agreements and can constitute trade secret misappropriation under the federal Defend Trade Secrets Act of 2016, exposing you to civil liability. Generate every answer entirely from your general technical knowledge, never from employer files.
• Maintain clean hardware boundaries. Complete all evaluation work on your personally owned computer, connected to your private home network, during off-duty hours. Using a company-provided laptop or working while clocked into your primary role voids statutory protections and violates company policies.
• Manage direct conflicts of interest. If your primary employer develops foundation models, fine-tunes large language models, or provides dedicated AI infrastructure to model developers, training third-party models may directly relate to your employer's business. Read your employer's code of conduct and outside-employment policies carefully. When in doubt, seek written clearance from your compliance team.

What to do this month

If you want to establish an engineering side hustle without jeopardizing your job, follow these steps over the next four weeks:

1. Locate and read your employment contract. Pull out your signed offer letter, employee handbook, and PIIA. Look specifically for invention assignment language, moonlighting restrictions, and conflict-of-interest definitions.
2. Check your state statutes. If you live in California, New York, Illinois, Delaware, Kansas, Minnesota, New Jersey, North Carolina, Utah, or Washington, confirm that your contract respects statutory exemptions for personal, off-duty work.
3. Audit your hardware and software setup. Ensure your personal laptop has the necessary development environments, CAD software, and secure browsers installed. Keep company hardware completely separate.
4. Practice with an AI interviewer. Test your spoken technical communication using our mock interview tool or Mercor's practice room to get comfortable answering adaptive technical prompts.
5. Apply to one AI training platform. Start by submitting an application to micro1 or Mercor. Take the proctored assessments in a quiet space with an external monitor disconnected and all browser extensions disabled.
6. Track your billable hours and tax obligations. Set aside a share of every payment for quarterly estimated self-employment tax in a dedicated account.
7. Expand across multiple platforms. Once approved on your first platform, apply to Alignerr to build task redundancy and secure consistent project volume.

The short version

Most engineer side hustles carry hidden costs: building apps risks employer invention assignment claims, while freelancing hits professional licensing walls and low hourly rates. Selling your technical judgment through AI evaluation avoids the ownership question and the PE gate, though the conflict and confidentiality lines still apply. Engineering roles outside software post typical ceilings of $90 to $100 and tops of $110 to $130 on micro1 and Mercor, and software roles reach $200 to $280 at the top. Explore live roles and start evaluating models on Skillora's AI training job board.