▶ Watch ↗20:45
Conference affiliation: Blackstone
John Stecher is Blackstone’s chief technology officer and a senior managing director. He leads technology across the firm, advises its investment teams and works with portfolio companies on technology decisions. His career spans enterprise software and financial products: leading work on IBM’s WebSphere, helping launch Goldman Sachs’s Marcus consumer finance business and directing technology at Barclays before joining Blackstone.
John began his career at IBM, where he led application development, architecture and strategy for WebSphere. IBM named him a Master Inventor, and he has more than 45 granted patents. By 2007, he led WebSphere Application Server’s Java enterprise performance teams and had spent five years working on SPECjAppServer as a lead developer and architect. His joint interview with IBM engineer Andrew Spyker discusses that work on benchmarking enterprise transaction-processing performance and scalability.
He spent more than seven years at Goldman Sachs in engineering and senior technology roles across capital markets, technology and consumer finance. As a member of the Marcus executive team, he helped take the consumer finance business from its initial concept and first line of code to its first customer. In August 2017, he moved to Barclays, where he served as group chief technology officer and group chief innovation officer, leading technology and product specialists across business units. He joined Blackstone as CTO in February 2020.
John holds a bachelor’s degree in computer science from the University of Wisconsin–Madison and a master’s in computer science from the University of Minnesota.
John credits his former Goldman colleague Marty Chavez with influencing his build-versus-buy decisions. He reserves internal engineering effort for proprietary capabilities that create value for the firm. Shared infrastructure is a candidate for an outside supplier; an unmet need across an industry can present an opportunity for strategic investment. His account of this approach connects his responsibilities for Blackstone’s engineering organization with his work evaluating technology companies.
Blackstone’s Secure Chat gives employees access to language models while protecting firm information. Its document-analysis tools help analysts search, extract and summarize material, linking results to the underlying passages so analysts can inspect the material behind an answer. These are products of Blackstone’s teams, whose work John describes as reducing routine analysis.
When evaluating software businesses, John asks what remains valuable as software becomes cheaper to create. He emphasizes proprietary data, knowledge embedded in workflows and customer trust. Software that customers depend on to operate their businesses has advantages that an interface over readily available data may lack.
John’s approach to agents starts with understanding the work to be done. He has used Factory and other coding tools himself to understand workflows and automate predictable steps in code, adding agents where routine decisions previously required a person. He calls this straight-through processing: agents supplement the automation of a workflow rather than replacing the need to design it. His engineering team has also used agents to write functional and performance tests and upgrade application dependencies, work he credits with increasing delivery speed.
Two engineering priorities support that approach:
Faster development also changes his investment decisions. John has deferred projects when the business did not urgently need them and a stronger model might make them cheaper to build in six months. But waiting indefinitely for the next model carries its own cost. He encourages engineers to build usable proofs of concept quickly, put them in front of colleagues and discard them when they do not prove useful.
He applies the same reasoning to replaceable internal applications: would rebuilding produce more value at lower cost than continuing to update the existing software? Earlier in his career, the time and expense of development made a rewrite difficult to justify. Cheaper implementation makes that question worth reopening more often.
A clickable prototype still leaves substantial production work. John points to deployment within enterprise infrastructure, cybersecurity review, supportability and software that can be traced and monitored through the firm’s operational systems. Agents have made it easier for him to return to hands-on engineering, but he still treats those production requirements as work that must be done. The appeal is personal as well as organizational: after years in senior roles, he can more readily turn his accumulated system-design knowledge into working software.
▶ Watch ↗20:45
Loading more talks and highlights…