Project Infinity, launched in May 2017 by NEX Group (formerly ICAP), could be the biggest bloodletting the distributed ledger sector has seen to date, having cost roughly $31.7 million and hemorrhaged dozens of jobs, former employees say.
Spearheaded by Jenny Knott, former head of NEX Optimisation, the project had a grand vision of bringing the company’s entire post-trade services portfolio onto one interoperable blockchain architecture. However, NEX issued a profit warning in October 2017 and just two weeks later Knott left the company.
In a blow for the nascent project, the results statement specifically said the investment in blockchain had reduced margins by 4 percentage points.
Sources say a deal to sell the business to CME for $5.5 billion, announced by NEX boss Michael Spencer in March 2018, was a driver in the cost cutting at the firm.
A former NEX executive, who spoke on condition of anonymity, told CoinDesk: “They scaled it back significantly. Spencer was not going to invest another penny that wasn’t going to get him short-term revenue.”
A blockchain industry source who said his company had been hiring former staff from Infinity, went further, alleging the project had not just been scaled back but had actually been “canned.”
“They got rid of Knott and then they quietly canned the project about two months ago; 47 out of 50 staff were fired,” said the source.
CoinDesk spoke to five former NEX employees, who all confirmed the extent of severe staffing cuts, if not the specific figures. One reported that, in his particular division, nine out of 10 jobs were lost, adding that there were a few divisions like his; another affirmed the figure.
“The people doing the strategic work mostly all disappeared,” he said.
In terms of the spiralling cost of the project, one former employee said NEX made a mistake when it decided to make Infinity a “program,” which brought added costs and expectations.
“When you make something into a ‘program’ it means hiring all these people who plan things and do tons of spreadsheets and so on,” said the former employee.
“There was an enormous group of people doing business requirements, managing user groups for opinions. Some of that stuff is necessary, but not at the scale it was done,” he continued.
Another cost factor related to Traiana, the division of NEX most closely linked to the project, they said, which was doing a upgrade of its infrastructure that ended up on the project’s books, thus increasing perceived costs.
Still, NEX itself is disputing the claims of former staffers.
Andres Choussy, CEO of Traiana, said Project Infinity had absolutely not been shelved. On the contrary, he said it has entered the execution stage in a live environment with blockchain builder Axoni. However, he admitted the project had been scaled back.
“Infinity had a vision; the vision is a sound vision which we still would love to try and get to one day. But we are being realistic,” said Choussy.
“You see this from a lot of the DLT initiatives, moving the market into something completely new is not an easy task, so what we have done is focused our efforts into specific applications, specific use cases that really are plan-driven,” he said.
NEX could not comment on the number of jobs shed at the project, but confirmed the number of staff involved had been reduced. No specific numbers were provided by the company.
Choussy said the DLT part of the project had been narrowed down to focus on post-trade FX.
Indeed, NEX was a returning investor in the recently announced Series A funding round in Axoni, which is said to be building “a massive FX post-trade data network.”
NEX is also an investor in DLT provider Digital Asset and was one of the early members of the Utility Settlement Coin (USC) project, although a source involved in USC said they have fallen out of communication with the group.
A former engineer on the project painted a picture of the tough challenge to meet the use case requirements with DLT, followed by a missed opportunity to get a product out the door.
Axoni was selected at the outset because it provided a unique partitioning capability that few of the other vendors out of the box did, said the former NEX engineer. Throughput had been an issue to begin with but transaction processing speeded up as the project evolved, he said.
In addition to the work done with Axoni, the engineer said a Hyperledger implementation was built which achieved privacy partitioning by creating different networks, as well as an internally constructed ledger, “basically a glorified replication scheme, but with cryptographic proofs and enforced privacy controls” which could handle very high throughputs.
“We were really right at the razor’s edge of getting this stuff out the door when there was a just a massive set of layoffs and the entire team in London was let go,” he said, adding:
“The politics around the situation changed, also the company was preparing to try to make things look good for the CME acquisition.”
Infinity’s DLT architecture was to anchor and immortalize an extremely rich data set coming from multiple parties – brokers, banks, buy-side – which would serve as a backbone for a one-stop financial cloud offering machine learning and AI as a service on top.
In the end, this became a stinging irony. An anecdote recounted by one former employee told how the day after the entire machine learning team was laid off, Spencer returned from Davos enthused about machine learning and AI and asking what NEX was doing in that area.
A proposal for a firm-wide AI strategy was even put together, said the ex-staffer.
Opinion seems to be divided about how well the remains of Project Infinity’s blockchain might fair once NEX becomes the property of U.S. exchange giant CME. The former NEX technology team member said the original vision of an all-purpose post-trade ledger with data analytics platform still has great value and may even be the “crown jewels” in CME’s eyes.
However, the former executive was doubtful about Infinity’s progress, saying, “CME have got their own distributed ledger strategy.”