This page has two parts. The first reports what the CAD and CAM listings on micro1’s portal say the work is, from their full descriptions. The second follows one invented part — a bracket — from the spec to the grade, as practice for one of those kinds of work: checking a parametric model against its brief.
The bracket is the blog’s own example. What the listings say the work is comes first.
Opens micro1. You apply and are vetted by micro1, not by this site.
What the CAD listings describe
On the date in Sources the full descriptions fell into four kinds of work.
Parametric modeling to a brief. Most of the listings describe making models: creating fully parametric 3D models of parts and assemblies from written briefs, editing existing models to new requirements, and auditing models for dimensional accuracy, tolerance application and manufacturability. They ask for GD&T to a named standard, design-for-manufacture experience across processes such as injection molding, sheet metal, CNC machining and casting, and engineering drawings. One describes building a library of parts and assemblies for an AI dataset. The listings titled for SolidWorks, Fusion 360 and Inventor specialists carry a note that matters: the project itself uses FreeCAD, and the named package is accepted as transferable experience. Python scripting for automation is listed as optional or preferred.
Writing questions about assemblies. Four listings describe a different job: studying multi-part assemblies supplied in native FreeCAD format and writing objective, non-guessable questions about them, each with an exact, verifiable answer — geometry, how the parts relate, what the assembly does.
Judging manufacturing guidance. Three listings describe giving expert feedback on manufacturing processes in CAM, CNC and Fusion 360 workflows, reviewing AI-generated content and recommendations, and developing the rubrics used to evaluate them. One of the three states that it is a coding-heavy role and expects experience with coding agents and Python.
Machining and CAM programming. One listing describes setting up and operating CNC machines — desktop and industrial — running first pieces and inspecting parts against drawings; another describes programming multi-axis toolpaths in Fusion 360, configuring post-processors and writing setup sheets and instructional material.
So “evaluation” describes only part of this category. Much of it is producing models, questions, toolpaths or rubrics.
The spec: a bracket that has to flex
The examples below are illustrative exercises created for this guide, not published micro1 assignments or interview questions. They show the kind of judgment the work needs; they are not a preview of what a project will hand you.
Which work this prepares for. The bracket case is practice for parametric modeling to a brief — specifically the auditing part: checking a model for dimensional accuracy, tolerances and manufacturability. A short exercise further down is practice for writing questions about assemblies. Nothing on this page practices CAM programming or operating a machine, and the case touches manufacturing guidance only through one finding.
The case starts with a task written by the expert, not the model. In this illustration it asks for a mounting bracket, in millimeters, for 3-axis milling with a stated end mill: a base plate with two slotted holes, a boss with a bore that takes a shaft, and a pocket on the underside to save weight. The spec gives the shaft diameter and the clearance the bore must leave around it. Three dimensions drive the part: overall length, bore diameter and pocket depth. The spec states the design range for each, and one relation: the slots stay centered on the boss when the length changes.
The pass criteria are written before any model runs: the part rebuilds across the stated range, the stated relation holds across it, every feature can be cut with the stated tool, and the bore leaves the stated clearance.
What came back
A complete feature tree, whether as native features or as a script that builds them: sketch, pad, boss, bore, slots, pocket, fillets. It opens, it renders, and at the authored values the numbers shown on screen are the numbers in the spec.
Read one: the render
A review that stops here passes it. The part looks like the drawing, the numbers on screen are the numbers in the spec, and the file exports. None of the four pass criteria has been checked yet.
Read two: sweep the dimensions, then read the model against the spec
Change the three driving dimensions across their range and record what breaks. In this invented case the sweep exposes the first two findings; the other three come from reading the model against the spec, line by line. These are stipulated findings in the fictional model, not claims about how every FreeCAD model behaves when a dimension changes.
- It does not rebuild. At other lengths the slot feature fails to rebuild: it depends on a reference that does not survive the change.
- The stated relation is lost. The base sketch is held by fixed coordinates, with no relation tying the slots to the boss. It is right at the authored dimensions; at other lengths the slots are no longer centered.
- A pocket the stated tool cannot cut. The pocket is modeled with sharp internal corners. A CNC manufacturer publishes, in its design guide cited in Sources, that a cylindrical end mill leaves a radius in an internal corner — a fabricator’s guidance, not a standard. On that basis, the pocket as modeled cannot be produced with the tool the spec names.
- Clearance missing. The spec states a clearance between shaft and bore; the model has both at the same diameter.
- A units slip. The fillet radius carries the spec’s number in inches, while the spec and the rest of the part are in millimeters.
Each finding is written with its cause: which reference the slots should have been positioned from, which relation expresses the centering, what corner radius the stated tool leaves. A note that says only “fragile tree” gives the next person nothing to act on.
Grading the part
The grid below is this blog’s own practice aid, not micro1’s rubric or a hiring criterion. Each row tests one observable thing. Its levels are cumulative: Strong is Acceptable plus the one thing named after “also”, and a row’s level is the highest one the part meets in full — a part that meets the “also” and fails the base requirement is at Fail. A row you have not checked, such as Rebuild before the sweep, is at Fail until you check it. A part with any row at Fail goes back, with the correction its Fail cell names.
| Criterion | Fail | Acceptable | Strong |
|---|---|---|---|
| Rebuild | A value inside the stated range produces a rebuild error. Correct the reference the failing feature depends on | Every value inside the range rebuilds without error | Also: a value outside the range is refused or flagged, not silently built |
| Design intent | A relation the spec states is lost at some value in the range. Correct it with an explicit constraint | Every relation the spec states holds across the range | Also: each relation is an explicit constraint in the model, not a coincidence of the numbers |
| Manufacturability | A feature cannot be produced with the stated tool or machine. Correct that feature | Every feature can be produced with the stated tool and machine | Also: the feature that needs the smallest tool is identified |
| Fit | A clearance or fit the spec states is absent from the model, or is there at another value. Correct the dimension | Every stated clearance or fit is in the model at the stated value | Also: its tolerance is stated, or the spec’s silence on tolerance is raised as a question |
| Units and values | A dimension the spec gives differs from it in value or in unit. Correct that dimension | Every dimension the spec gives matches it in value and unit | Also: the unit system is declared in the file |
The five findings map onto the five rows, one each: the slot feature that errors is a Rebuild fail; the lost centering is a Design intent fail; the sharp-cornered pocket is a Manufacturability fail; the missing clearance is a Fit fail; the fillet in inches is a Units and values fail. The generated bracket fails all five rows and goes back — with five corrections, each to one feature. None of them calls for remodeling the part from scratch.
A second exercise: questions about an assembly
Practice for writing questions about assemblies. Model a small assembly of your own — the bracket, a shaft and two fasteners will do — and write three questions about it, each with one exact answer that can be checked in the file: a distance between two named faces, which part constrains which, what happens to one part when another is removed. Then try to answer each question without opening the file. A question you can answer by guessing from the part names is not finished.
The assembly is yours, built for the exercise. Do not use an employer’s or a customer’s files.
Who the case is written for
Mechanical designers, drafters, manufacturing engineers and machinists who model parametrically. The descriptions name specific tools, so say which ones you use. Useful profile lines:
- The package and version family, and whether you use it professionally or self-taught.
- FreeCAD, if you have it. Several descriptions say the project runs on it.
- Years of parametric modeling. Some descriptions set different thresholds for editing work and for creating complex models.
- GD&T and the processes you have designed for, named: molding, sheet metal, machining, casting.
- CAM and machining experience separately from design — CAM and CNC listings are their own roles.
- Scripting, if you have it. Python appears as optional in the modeling listings and as central in one manufacturing-evaluation listing.
Interview preparation from the same bracket
An AI interview — step two of the four micro1 lists, between applying and onboarding — and, per micro1, human recruiters who review profiles and match experts against project requirements.
Suggestions from this guide. Keep two things apart: your professional experience, which has to be true and limited to what you may share, and a practice part, presented as practice.
- Narrate a modeling decision on a practice part. The bracket above, or one you modeled for the purpose: the reference scheme and feature order you chose and why. Say it is a practice part.
- Describe a design change you made for manufacturing reasons, at the level you may share: the process and the kind of change, not the customer’s part or its dimensions.
- State your tools precisely, including the ones you would not claim.
- Say how far your tolerancing goes. Whether you work to GD&T callouts or to plain dimensions; the descriptions that ask for GD&T name a standard.
Conditions, IP and eligibility
Do not submit employer or client geometry. Model hypotheticals from scratch, as the bracket above was.
On pay, the card and the description have to be read together. The cards on the opportunities page show an hourly range. The three manufacturing-evaluation descriptions then state:
Compensation is output-based; experts are paid per task that meets the project specifications. The time required to complete work may vary depending on the expert’s experience and workflow. Minimum submission requirements apply. Experts must submit a minimum of tasks per week.
One of them also gives a weekly schedule in hours, next to that per-task wording. The modeling, question-writing and machining descriptions say nothing about how pay is calculated, so for those the unit of payment is unknown. In neither case is there a price per task, a number for the weekly minimum, or a statement of what happens to a model or rubric that is rejected.
Where a description states per-task pay, a task that runs long — a rubric that needs a second pass, a review that takes more reading than planned — may lower what the work pays per hour in effect; how that extra time is treated is for the contract to say. Before accepting, ask in writing what a task pays, who accepts it and against what, whether rejected work can be reworked, and what the weekly minimum is — and do not treat the hourly range as payment for each hour at the desk until the contract says so.
No rate is printed here, by this page’s own choice. On a separate point, micro1’s referral terms forbid anyone promoting the program from making promises about compensation, placement or acceptance.
Whether you need your own seat of commercial software depends on the listing; the modeling descriptions point to FreeCAD, which its own site describes as open-source.
micro1 states it does not hire residents of: Afghanistan, Belarus, China, Cuba, Democratic Republic of the Congo, Hong Kong, Iran, Iraq, Libya, Macao, Myanmar, North Korea, the Russian Federation, Somalia, South Sudan, Sudan, Syria, Venezuela, Ukraine and Yemen. Snapshot from the date in Sources.
Do referred designers get assessed differently? Nothing we found says they do. The vetting belongs to micro1, not to this site, and micro1’s referral terms rule out promises of placement, compensation or acceptance from anyone promoting the program. A page promising that a referral helps is a warning sign.
What to have ready: one practice part that survives a dimension sweep, three questions about an assembly you built, a truthful account of your manufacturing experience at the level you may share, and a precise tool list.