Know a Tool, Know Federal Work, or Lead Engineering
The November 4 saved records show several distinct routes into AI work. Fourteen hourly roles focus on skill in a single program, such as KiCad, FreeCAD, LTspice, Godot or Krita. Eight are built on US federal government work, from grants and budgets to criminal investigation. Eight more want visual creatives: photographers, set designers, stylists and editors. At the top, three engineering leadership roles list up to $300/hr.
93
eligible records in this window
10
platforms in the same window
What's in this drop
Know one program well
Fourteen saved hourly roles are each built around one program. You do real work in it, such as laying out a circuit board in KiCad, modelling a building in FreeCAD or setting up a scene in a game engine, and explain every step so an AI system can learn the workflow; most also prefer a relevant master’s degree or PhD. KiCad, FreeCAD, LTspice, NgSpice and Qucs-S list $65 to $123/hr, LibreCAD has the highest top rate at $65 to $165/hr, and 3D Slicer for medical imaging lists $60 to $112/hr. The game engines (Godot, Solar2D, Open 3D Engine) list $70 to $130/hr and Panda3D $50 to $120/hr; a separate Godot Game Developer role pays per accepted task, with no supported numeric rate in the saved source. The drawing and pixel-art programs (Krita, LibreSprite, Inkscape) start at $21 to $25/hr and top out at $70/hr. See STEM and coding.
Federal work, as a skill
Eight roles ask for experience of how the US federal government actually works. Grants specialists draft funding announcements and score applications; budget analysts are asked for federal budget experience at about the GS-11/12 level; IT specialists for hands-on federal IT work; veterans service representatives review AI-written disability claims. Rates run from $50 to $70/hr for regulatory policy and $60 to $85/hr for grants, budgets and veterans claims, up to $70 to $100/hr for criminal investigation and $80 to $105/hr for IT. A student loan specialist role asks for a year advising borrowers on federal loans, at $30 to $50/hr. Alongside them, three medical coding and chart-review roles for people in the US want CCS, CIC or CPC certification (the chart reviewer also takes an RN or clinical background), at $50 to $70/hr. Start with the Medical Records Coder listing or browse medical.
Visual creatives, and the top of the range
Eight creative roles cover photography and cinematography, set and production design, fashion and interior styling, graphic and editorial design, brand and social creative, art buying, curating and photo editing. Their saved pay headers and descriptions disagree, so this review does not quote a rate for them. Confirm pay with the platform before applying. At the top of the advertised pay range, a chief engineer or senior mechanical engineer role lists $200 to $300/hr, and a systems architect role and a lead systems engineer role list $150 to $300/hr; all three prefer US-based people and also consider others with strong experience. An engineering manager role for people in the US with 10+ years of experience is also in the saved records, and an open-source maintainer role lists $150/hr for 40 hours per week on weekdays, on an exclusive basis. That role requires authorization to work in the United States on a W-2 basis.
Language work and paid studies
Music producers and audio engineers who work in English, French, German, Japanese or Korean can rate AI-generated songs; pay is per task in some languages and hourly in others, so read each listing. The saved records also include search-rating roles in Italy and Germany, data roles in Tamil and Telugu in India, and medical translation roles in Farsi and Thai for translators based in the US. Then the paid studies, which state a fixed or per-submission payment instead of an hourly rate: a shopping-experience study of about 5.5 hours for native speakers in France ($130), Germany ($182), Japan ($143) and the UAE ($195); a three-day study in Midtown Manhattan for native speakers of several languages at $1,200 to $1,600; and a one-time phone fitness-data contribution in the US that most people finish in under 30 minutes. An ongoing video-recording project in India and Vietnam lists $4 per accepted hour of video, up to 60 hours each. See data collection and translation.
Up to ten eligible listings, ordered by first observation
Benefit Navigation expert
Check whether AI benefits guidance reflects real coverage decisions
Loan Underwriter / Loan Officer — Paid Research Survey (US)
Share lending judgment in paid Stanford survey research
Systems Architect / Chief Systems Architect
Design systems architecture challenges for AI training
Chief Engineer / Senior Mechanical Engineer
Build realistic mechanical engineering projects for AI evaluation
Principal / Lead Systems Engineer
Define systems engineering challenges that test AI reasoning
Senior Protection Engineer (Medium Voltage / Utility Distribution)
Evaluate medium voltage protection engineering for AI training
Materials Scientist (Nanoparticle Sizing & Lot Release)
Assess nanoparticle sizing and lot release through manual analysis
Engineering PhD
Build verifiable engineering problems for a short research pilot
Veterans Service Representative
Evaluate veterans benefits decisions to improve AI training
Federal Contract Specialist
Create realistic federal acquisition cases for AI evaluation
Methodology
Editorial examples use saved November 4 records. The counters and cards show records still eligible in the current catalogue, first observed from November 1 at 23:59:59 UTC through November 4 at 23:59:59 UTC. November 4 collection coverage is partial. First observed is not the employer publication date, and these changing counts are not an archived total. The cards include up to ten records with stated pay; the total also includes records without stated pay. Story amounts are advertised US dollars, with hourly, per-task and one-time payments distinguished; they are not guaranteed earnings or paid hours. Pay conflicts are omitted. Check each listing for its current availability, location and qualifications. Apply buttons retain the platform application links. Some links earn SignalGigs a referral fee.
For more listings, browse the current collection. For a previous weekly recap, see Friday Flash.
Frequently asked questions
What experience appears in the saved November 4 AI-work listings?
The saved records include roles based on skill in one program, US federal work, visual creative work and senior engineering. Three engineering leadership listings advertise hourly ceilings of $300. These are advertised amounts, not promised earnings. First observation dates are not employer posting dates.
What does a single-program expert role involve?
You use one program to complete real tasks, for example designing a circuit board in KiCad or building a scene in a game engine, and explain each step as you go. The listings ask for experienced users of that program. Listed pay runs from $21 to $70/hr for drawing and pixel-art programs, with engineering and CAD examples at $65 to $165/hr. Each amount is an advertised range in US dollars; eligibility and available work vary.
Which of these roles are limited to the United States?
Several are. The three medical coding and chart-review roles, the federal contract specialist, student loan specialist and regulatory policy analyst roles, the engineering manager role, the English voice, digital artist and visual content evaluator roles, the Farsi and Thai medical translators, a data quality rater role, the phone fitness-data study and the loan underwriter survey all ask for people in the United States. A Brazilian Portuguese content reviewer role is posted in the United States, a maps testing role is based in Austin, Texas (remote, on Central-time shifts), and the three-day study is in New York. The three engineering leadership roles prefer US-based people but also consider others with strong experience. Location limits vary by listing; check each one before you apply.
What do the dates and counts in this Wire mean?
The date is when SignalGigs first observed a listing, not when its employer posted it. Counts and cards use currently eligible saved records in a fixed 72-hour window ending November 4, 2026 (UTC). November 4 collection coverage is partial. Counts can change as listings close or age out, so this is not an archived census of all jobs originally available.