Why pre-screen smart glass developers before the technical panel
These devices get hot, throttle and run out of battery faster than any demonstration suggests. Anything that runs continuous computer vision will hit a thermal limit within minutes, and the user cannot put the device down without stopping the task. Developers worth hiring have watched that happen and designed around it. A short screen asks what happened after ten minutes of continuous use.
What actually matters when screening Smart Glass Developer candidates
- 01
Technical depth
Check depth in switchable glazing stacks: electrochromic tungsten oxide layers, PDLC and SPD films, sputtered low-E coatings, busbar design, drive voltage curves and switching kinetics.
- 02
Work that shipped
Probe products that reached glazing lines: laminated IGU builds, ASTM E2141 durability cycles, EN 1279 seal testing, façade pilots, yield per square metre and field failure rates.
- 03
Diagnosis under uncertainty
Test how they chased defects: edge seal delamination, iris effect, non-uniform tinting, busbar corrosion, moisture ingress, using EIS, spectrophotometry and accelerated weathering data.
- 04
Working across the org
Assess work with façade consultants, architects, controls integrators and BMS or KNX teams, plus coordination with laminators, tempering lines and building code reviewers.
Pre-screening questions to ask Smart Glass Developer candidates
12 questions grouped by what they test. Ask the same set in every screen and score answers on a consistent scale, or send them as an async video screen and compare answers side by side.
Ran on real devices
3 questions01Can you describe your experience developing software for head-worn devices?
Listen forApplications that ran on real hardware with users, and the session lengths they were actually used for.
Development done in simulators, or applications only demonstrated for a few minutes at a time.
02Have you developed applications for specific industries using these devices?
Listen forA real use case such as field maintenance or warehouse picking, with the workflow properly understood.
Use cases described speculatively, or applications built without observing the work they support.
03Which hardware platforms have you worked with?
Listen forDevices used at length, with an honest account of each platform's display, field of view and battery limits.
Platforms named from specifications, or limitations described only from marketing material.
Thermal and battery
3 questions04How do you approach optimising performance on these devices?
Listen forThermal throttling and battery drain measured during sustained use, with processing offloaded where sensible.
Performance measured in short bursts, or thermal behaviour never observed over a long session.
05Can you discuss a project where you used computer vision on these devices?
Listen forVision run at a duty cycle the device can sustain, with accuracy measured under real lighting conditions.
Continuous vision assumed feasible, or accuracy measured only in controlled lighting.
06Have you worked with spatial mapping or environmental understanding?
Listen forTracking limitations understood in real environments, including featureless walls and changing light.
Tracking assumed reliable, or environments that break it never encountered in testing.
Hands-free interaction
3 questions07What experience do you have with gesture or voice input on these devices?
Listen forInput designed for occupied hands and noisy environments, with a fallback when recognition fails.
Voice assumed to work in a factory, or no fallback when an input method fails repeatedly.
08How do you ensure an interface is usable on a head-worn display?
Listen forMinimal information in view, positioned so it does not obstruct the task the person is doing.
Phone interface patterns transferred directly, or content placed centrally in the user's field of view.
09Can you give an example of addressing a user experience problem on these devices?
Listen forA problem found by observing real use, such as fatigue or obstruction, with the design change described.
Problems identified by the team only, or no observation of extended use in the actual environment.
Camera privacy
3 questions10How do you handle privacy and security in applications for these devices?
Listen forBystander privacy considered, with recording indicated clearly and capture limited to what is required.
Continuous capture with no indicator, or bystander consent not considered in workplace deployments.
11What steps do you take to test and debug applications on these devices?
Listen forTesting during long sessions on target hardware, with logging that works without a tethered connection.
Testing done tethered to a workstation, or no way to diagnose a problem during real use.
12Can you explain your experience integrating these devices with other systems?
Listen forIntegration with existing work systems, with connectivity loss handled so the worker is not blocked.
Constant connectivity assumed, or the device unusable when the backend is unreachable.
How to score responses
Score every candidate on the same four criteria immediately after the screen. At this stage you are shortlisting for panel interviews, not making the final call.
Technical depth
35%5Explains EC layer chemistry, ion transport limits and drive electronics, with measured switching times, haze and visible transmittance ranges.
Work that shipped
30%5Names shipped panel sizes, production yields and installed projects, including warranty returns and the design change that reduced them.
Diagnosis under uncertainty
20%5Walks through a real defect from field complaint to root cause, citing EIS or cross-section evidence rather than guesswork.
Working across the org
15%5Describes translating architect tint targets into controller logic and manufacturing tolerances agreed with the laminator and glazing contractor.
Continuous vision hits a thermal limit in minutes and the user cannot put the device down. A one-way video screen asks about that.
Try it on HirevireScreening FAQ
Process basics
How long should a pre-screening round for this role take?
Fifteen minutes across eight to ten questions, answered async. Enough to establish what ran on real hardware, test their performance thinking, and check interaction and privacy handling.
Should I expect production deployments?
In industrial settings, sometimes. Consumer deployments are rare. What matters is whether they built something that ran on a device for a sustained period with real users.
Evaluating answers
What is the strongest signal when screening this role?
Knowing what happens after ten minutes of use. Developers with device experience describe thermal throttling and battery drain. Anyone whose demonstrations were brief has not hit the limits.
How do I judge their interaction design?
Ask how they handle input when hands are occupied. Real answers cover voice, gesture and their failure modes in noise. Anyone porting touch patterns has not designed for head-worn use.
Go deeper on this role
Sanat has been hiring since 2012 and watching the recruitment industry change up close ever since, and turned that screening process into Hirevire's video screening platform. LinkedIn
