How Deployment Flexibility and Model Choice Shaped Blackstone’s Software Factory Strategy

Blackstone was looking to build a software factory that could operate within its existing technology environment, under its security controls and direct observation. Factory gave the firm that deployment flexibility without limiting its choice of models or engineering workflows.

“It was absolutely critical to ensure that agents ran inside of our data center, under our control, underneath our observation.”

John Stecher, CTO at Blackstone

Watch the video to see how Blackstone is putting that vision into practice.

Blackstone is the world's largest alternative asset manager, serving institutional and individual investors by building strong businesses that deliver lasting value.

The Technology organization at Blackstone oversees technology across the firm and advises its investment teams. It also provides technology guidance to a portfolio spanning more than 270 companies.

Blackstone's goal for AI is practical: make its people more effective and efficient while maintaining security and enterprise controls. They trialed a number of offerings on the market and concluded Factory was the only one that fully met their requirements.

"I think every single engineering leader out there today - I think every single business, in general, out there today - should be thinking about how to build a software factory in their business to out-innovate their peers in the market. Our interest in Factory was not just for software development. It was really thinking about my whole organization and what we do on a day-to-day basis." John Stecher, CTO at Blackstone

Blackstone is building its software factory.

Why Factory

When Blackstone evaluated what it needed from a next-generation software engineering platform, one requirement was critical: the system had to support every phase of the software development lifecycle while fitting within the firm’s existing operating environment and controls.

"A lot of the tools out there are very targeted… addressing code reviews, or building a feature," says Ines Sheppard, Managing Director, Blackstone Technology and Innovations. "What's really nice about Factory is that it thinks about the full lifecycle of software development."

Two key requirements mattered in particular: where the agents would run and whether Blackstone could retain its choice of models.

Deployment flexibility

Blackstone needed agents to run within its existing technology environment. Factory supported deployment inside Blackstone's data center, a critical requirement for the firm.

"Remote deployment and remote engineering were the key things that we looked at when we were evaluating Factory," Stecher says. "It was absolutely critical to ensure that agents ran inside of our data center, under our control, underneath our observation."

Factory’s core Droid runtime can run in Factory’s cloud computing environments, customer’s on premise infrastructure, or a fully air-gapped environment. That deployment flexibility lets enterprises keep execution within the boundary they choose. For Blackstone, it meant Droids could run under its control and observation.

Model choice without provider lock-in

Blackstone also wanted flexibility across model providers while working within its existing governance and security controls.

"Model choice was a big one," says Monty Hall, CTO of Data, Cloud and Developer Experience, Blackstone. "We didn't want lock-in with any single lab or provider and thus constrain innovation. We also have very strict controls around governance and security."

Factory is model-independent by design. Customers can use approved models through Factory's managed catalog, an enterprise gateway or self-hosted deployments. That matched Blackstone’s need to retain model choice within its existing governance and security controls.

What changed with Factory

With Factory, engineering teams at Blackstone use Droids across the entire SDLC, including code review, testing, documentation, and larger engineering tasks. Shared context across these workflows helps developers spend less time rebuilding information between steps.

Factory is also reducing onboarding friction at Blackstone. Using Factory and dev containers, getting a new developer set up and contributing to the codebase has gone from up to two weeks to minutes.

"Today, developers are able to avoid the friction of onboarding, finding documentation… and focus on much higher-value work, like architecture and decisions around design," says Hall.

Blackstone’s engineering teams are also using Factory for larger units of work spanning planning and implementation.

On one project, a requirements document and working proof of concept that previously would have taken product managers and engineers about a month took one engineer roughly a week and a half.

Building Blackstone's software factory

After evaluating the market, Blackstone chose Factory for its deployment flexibility, model choice, and enterprise-first approach.

Stecher’s long-term goal is for Factory to become an extension of his team. “I see it augmenting my already-great people, making them better product engineers, software engineers, cybersecurity engineers, system reliability engineers by allowing them to partner up with Droids to deliver more for the firm.”

Stecher sees Factory as part of a broader shift in how companies build differentiated software. Building a software factory that autonomously improves repeatable tasks enables and empowers teams to build faster than ever before.

“I genuinely believe we’re in the next great era of software development, where a small team of engineers can produce a lot of software that differentiates your business from your competitors,” he says. “It’s what we’re aiming to do at Blackstone.”

Learn more about Factory for financial institutions.