FYI, if you make ASICs you likely also need to test them at one point. I am currently working on a test setup for a (non-Sky130) test tape-out I did mid last year. You can find first blurb on my blog.

Nice. I also bought an eval board for the AD5522 as it seemed like quite a good starting point for a reasonable price, and made some I-V curves of LEDs etc. to test it. (IIRC they tied force and sense together on the PCB but that could be fixed easily with a scalpel and some enmelled wire to some 4mm terminals.) I somewhat lost interest when I realised that the first job would be to write some better software for it as I'm not much good at that. Perhaps if I could interest Whitequark or the Sigrok developers in doing it....or you...... I think a new board that has a FPGA and AD5522 and some good ADCs, (and a quiet power supply isolated from USB and mains earths) could make a really useful quad SMU. Maybe adding an extra channel with an AD5560 or some other current boosting scheme would be good as well.

Yeah dev board seem to be designed to limit real life usability. Don't think a FPGA is needed; any MCU with SPI interface should work. If it has on-chip NVM it could store calibration parameters. Currently using a Raspberry pi to connect to the dev board and doing software development in Python. Plan is to drive measurement and plot results in a Jupyter notebook.