Licensed in Massachusetts to teach Math 8-12 and Physics 8-12, I specialized in preparing at-risk high-school students for the Massachusetts Math exam. My blog (see above) describes my experience and makes some suggestions.

I wrote and successfully defended US Patent 6272672, "Dataflow processing with events" (115 figures, 151 claims). It describes a new and universally applicable dataflow execution model for applications with bidirectional graphical user interfaces. The value proposition for the technology is at http://www.melconway.com/hand-eye/software-houses.pdf .

Working with executives from IBM and TERC on the development of IBM's graphical PC-based classroom science workstation, I united into one ten-person team software engineers from these two companies and managed the successful completion of the development project.

I worked with Wang engineers on the introduction of object-oriented technology into the Wang product line.

I independently developed and brought to the company the technology behind Mac Pascal and Apple II Instant Pascal. Mac Pascal was instrumental in helping the Macintosh enter the undergraduate computer science market.

Working with Monroe's engineers, I designed and programmed a minicomputer-based internal application development tool for a bookkeeping machine, did the logic design for a TTL-technology programmable calculator, and wrote the internal firmware for a cartridge-programmed vertical-market calculator.

Under contract to Massachusetts General Hospital I built the first freestanding digital video viewer for medical CAT scanners. MGH’s experience led to two more contracts, from the Washington University Medical School and Brookhaven National Laboratory, for PET scanners. All of these machines were entirely of my design and execution; except for outsourcing the construction of the memory boards which I had designed, I did everything: proposal, negotiation, all design, drafting, purchasing, construction, test, documentation, installation, training, and service. I delivered two radiology conference papers describing these machines.

Using the one-pass COBOL compiler design I had published, I delivered internal compiler design documents for the DEC PDP-11, the Burroughs 270, and the Honeywell 200. The latter two machines were medium-scale business data processors and the compilers were released as products.

The National Bureau of Standards engaged me to document the medical-content acquisition process of the PROMIS (Problem-oriented medical information system) project from a systems analysis perspective. PROMIS realized a vision of Lawrence L. Weed, M.D., a professor of medicine at the University of Vermont Medical School. More than a medical record system, PROMIS was a provider guidance system that maintained, in the patient's record, traceability of the provider's actions to the medical literature. (This aspect of health care quality control has been lost in the public conversation.) I was also engaged to specify a new institution that would maintain currency and distribute updated medical content in PROMIS systems. On another assignment I wrote the technical specification and first user manual of the MUMPS language as part of its documentation as an ANSI standard language. I also consulted to the Veterans Administration on adoption of MUMPS technology within the VA health-care system.