Dr Charles Swanton's days are busy, but the work is rewarding.
The Francis Crick Institute, where he works, has been testing medics in London who were sent home after showing symptoms of Covid-19.
Those found to be free of the virus are now back at their jobs.
Overburdened hospitals have been desperate for more testing facilities like this to help medical staff return to work.
The UK government set a target of 100,000 tests per day by the end of April, but has struggled to get near that number.
A test to see if someone has the coronavirus is a complicated process (different tests, which see if someone has ever had the virus are still awaiting approval).
Molecules on a swab are broken down into genetic code, using chemicals, liquid handling robots and a PCR (polymerase chain reaction) machine which can make billions of copies of DNA strands.
Private and university laboratories across the UK have donated enough equipment for three huge testing centres in Glasgow, Milton Keynes and Alderley Park in Cheshire.
But having the machines is not enough, they also need blended cocktails of chemicals to function.
These secret recipes have been tested over time, verified by regulators and guarded by the companies that sell them.
Like a cook with a ready-bake cake mix, scientists know all the ingredients, but the exact proportions are specific to each company.
The firms that manufacture and sell them include Qiagen, Roche, Merck and Eurofins Genomics. Each have their own recipes, designed for specific models of the PCR machines.
Dr Swanton and his colleagues at the Crick Institute realised that most of the world would be clamouring for these kits. They also knew the firms that manufactured them would be swamped.
So rather than wait, they reverse engineered their own "home-brew" to test local medical staff in London, as a voluntary service.
The Crick Institute is led by Nobel Prize-winning scientist, Sir Paul Nurse, but not run by the health services.
It is a research lab formed from a partnership between Cancer Research UK, and London hospitals which include the Royal Marsden, Imperial College London, King's College London and University College London.
Three weeks ago, when the virus crept across Europe their labs were deemed non-essential and closed. They handed in much of their machinery to the Department of Health and Social Care which is leading the testing ramp-up, outside hospitals.
Dr Swanton, at the time, worked as Cancer Research UK's chief clinician, researching the way that cancer progresses.
"We were going to be sent home. I thought to myself, 'Well there are a lot of non-essential workers I know who might actually be quite essential to the coronavirus effort,'" he says.
He sent an email round. A working group was formed. At the same time, Mr Nurse sent an email to his employees at the Crick Institute asking for possible volunteers for a lab.
He received 300 replies in 24 hours.
Clinicians including Dr Swanton, many from the University College London, and some at a private lab nearby in St Pancras called HSL (Health Service Laboratories), worked together to find a new procedure for making the chemical kits with the equipment that the government had left behind.
Dr Swanton now leads the Covid-19 testing efforts at the Crick Institute.
"We have the staff here, the facilities, the resources, the reagents and the know-how to get on and just do it. And so that's what we chose to do," says Dr Swanton.
Their home-made chemical kit (an RNA extraction kit) has been approved through a rapid accreditation process and they have made their operating procedure public.
How the chemicals work
The test for Covid-19 uses two stages of pre-packed chemical kits to extract the genetic material from the mucus and cells found on a skin swab.
RNA Extraction - around £350 for one pack of 50
In this part of the test, the virus's genetic code, its RNA, or ribonucleic acid, is found, cleaned and separated.
Enzymes and other chemicals break up all of the cells that are sent in on the tip of the swab.
Enzymes called proteases break up proteins in the sample. (Similar chemicals are added to laundry detergent to cut up the proteins in food stains.)
Another set of chemicals stick the RNA to a membrane.
And the reaction takes place in a liquid called a buffer.
At this point, the RNA is still too small to be detected. This is when the second chemical cocktail comes into play.
PCR kit - Around £250 for one pack of 200
This chemical cocktail allows the virus RNA code in the PCR machine to reproduce itself.
Here the RNA is converted to a form of DNA, covered in fluorescent chemicals and copies are made until there are enough to detect.
If enough bright spots show, the test indicates that the virus was present in the sample.
The companies that are better known for making these chemical kits warn against reverse engineering them under the current circumstances.
Dr Thomas Theuringer, a spokesperson for Qiagen, a German chemicals company that supplies reagents to the UK, says replacing these reagent cocktails with home-made recipes is "playing with fire".
"We can only guarantee that our extractions work if we make them in our production facilities where we have a controlled environment. Any mis-step and you might get a false positive and create more harm than good," he says.
Several reagents produced by the Centers for Disease Control (CDC) in the US failed to produce conclusive results. The CDC later admitted that kits had been "rushed".
"We are not talking about baking a cake - this is about life and death," says Dr Theuringer.
The benefit of using commercial solutions, he says, is that Qiagen has been making them for a long time and that the standard operating procedures in their labs have been verified by several international health organisations.
Roche, a company which also currently makes reagent kits for UK testing sites, agrees.
"The primary obstacles in another company or manufacturer producing any Roche test and reagents are time and expertise. Roche cannot guarantee safety and reliability if the reagents required for the test were manufactured outside our production network," a spokesman told the BBC in an email.
Stanford University professor Eric Kool says: "People have used home-brews for RNA extraction for a long time but for scaling up testing needs to be done in an automated fashion so that you can process many samples."
Prof Kool teaches chemistry at Stanford, has his own RNA extraction kit business and offered to help the effort in the United States, but testing centres there told him his kits were not the right kind for the machines they had in place.
"Kind of like printer ink - you have to buy the right one for the printer," he says.
"These kits are all corporate secrets," he says, but adds that the automated kits have plug-ins which work best with the kind of 24-hour testing that needs to be done with Covid-19.
"Even while people are sleeping, the robots can be running samples," he says.
This is the nub of the problem with large testing sites, like the ones built by the government.
The PCR machines they have collected from labs across the UK will work best with the chemical kits which are in the shortest supply.
The problem is worldwide, everyone is trying to get hold of the same automated RNA extraction kit.
To add to the complexity, the companies that sell them have differing ideas on which buyers should take priority.
This has been a difficult prospect for Qiagen, says its representative, Dr Theuringer, because his company has had to find a version of what fair allocation might be.
"No one company alone can help meet the demand," he says.
It went from manufacturing 1.5 million kits a month to 20 million per month, hired new staff and moved from a three-shift day to working around the clock.
Even so, they have been unable to meet demand, and even passed buyers on to their competitors.
Roche says it remains committed to its partnership with the UK government.
But its spokesman warns that "due to the high demand on reagents and consumables the supply situation may be challenging in the short term in some cases".
Dr Swanton cannot tell whether making their own test kits for local NHS staff was the right decision.
"Only time will tell," he says, "But we thought that doing nothing was not an option, really."