The software development life cycle has always pretended to be a straight line. Requirements in, code out, ship it, watch the graphs. Andrew Lau says that line is about to dissolve, a claim now circulating across tech news as AI reshapes engineering work, and he is not saying it from a podium or a pitch deck. He said it in a McKinsey interview published December 9, 2025, with the calm confidence of someone who has already watched one era end and another crawl out of the surf, a tone that has caught attention in recent tech news coverage.
Andrew Lau is the co-founder and CEO of Jellyfish, the Boston-based software engineering intelligence company founded in 2017 alongside Philip Braden and David Gourley. Jellyfish was born out of lived frustration, not theory. After years inside massive engineering orgs, including Andrew Lau’s time leading engineering at Endeca before its $1.075 billion Oracle acquisition, the team saw the same problem everywhere. Engineering was critical, expensive, and strategically opaque. Everyone felt it. No one could see it, a blind spot increasingly highlighted in tech news about enterprise software.
In the McKinsey conversation with Martin Harrysson and Prakhar Dixit, Andrew Lau drops the line that keeps echoing in tech news circles. He fundamentally believes the software development life cycle will be completely redefined within three years. Not tuned. Not optimized. Redefined. Code stops being the hard part. Intent becomes the constraint. Writing software gets cheap. Deciding what should exist, why it exists, and how it creates value becomes the human work, a reframing gaining traction across tech news focused on AI and productivity.
That is where Jellyfish lives. The company has grown to serve more than 500 enterprise customers from Boston, tracking engineering work as a living system instead of a pile of commits. Their platform pulls signals across the full lifecycle and ties them back to business outcomes, cost, and capacity. In December 2025, Jellyfish expanded its AI impact measurement across the entire SDLC, because measuring AI at the keyboard level misses the current, a product move now noted in tech news covering enterprise AI tools. The value moves through the whole body.
The metaphor fits. Jellyfish do not fight the ocean. They read it. They move with it. As AI agents creep into planning, testing, delivery, and observation, organizations that cannot see the flow will mistake motion for progress, a warning increasingly echoed in tech news commentary. Andrew Lau’s warning is not about tools. It is about measurement, roles, and the uncomfortable truth that many leaders do not know what their teams are actually building anymore.
This moment matters because it puts a clock on the wall, a sense of urgency now reflected in tech news about the future of software teams. Three years is not theoretical. It is budgeting cycles, org charts, promotion paths, and incentive models that were built for a different tide. Engineering leaders who still measure output instead of impact will feel this shift first, and not gently, a forecast tech news readers are beginning to take seriously.
