If there’s one word we’ll hear continue to echo throughout 2024, it's efficiency.
Between startups ditching the “growth at all cost” mentality for more sustainable milestones in today’s slow market, and customers doubling down on technologies that will drive business efficiencies through automation, there’s a realignment towards core business fundamentals. The staggering AI advancements happening everyday (shoutout to OpenAI) shows that there’s no shortage of new tech solutions and capabilities being brought to the table to meet these needs.
As per tradition, here are the top enterprise software predictions our Work-Bench team and portfolio founders foresee in the upcoming year:
On AI-Generated Code: While my earlier prediction that "code would increasingly be written and edited by code" came to fruition for most developers, the rapid adoption of AI coding assistants also highlighted the limitations of LLM-based tools. Looking ahead to 2024, I anticipate a significant shift driven by data providers with dynamic, runtime data and advanced code quality analysis capabilities who have the data key to surmounting these challenges. By moving beyond the current models with enhancements and context, we will transition from the 10% productivity improvements currently reported to much more exponential gains. Expect further accelerated development cycles, superior code quality, and enhanced architectural design, marking an unprecedented synergy of human expertise and AI efficiency in the software development domain.
Elizabeth Lawler, CEO & Co-Founder of AppMap
On Application Monitoring: The increase in AI-generated code, 3rd-party dependencies, and transient containerized workloads will make it more difficult for teams to reason about and troubleshoot application issues. At the same time, cloud cost concerns will reduce the ability of teams to keep throwing resources at performance problems. These trends will put more pressure on engineering teams and slow down feature development. Engineering leaders will take a fresh look at their application monitoring strategies, opting for more powerful solutions, as team velocity slows and observability costs balloon.
Lyndon Brown, Co-Founder and CEO of Prequel
On Open-Source LLMs: The AI/ML ecosystem will double down on open-source models, creating a free and open community for builders to work with LLMs and avoid vendor lock-in. While hobbyists will continue to leverage closed-source models, true builders will graduate to more customizable open-source models in order to build for unique use cases. On the flip side, we’ll see behemoth closed-source model provider platforms like Anthropic, OpenAI, etc. continue to invest in their own LLMOps and infrastructure tooling, killing the chances for stand alone LLMOps / infra startups to compete.
Danny Chesley, Investor at Work-Bench
On Data Quality Enforcement: As companies move to optimize spending in 2024, teams are going to be pushed to do more with less. One major challenge with most large companies is that the manual work involved in communicating and resolving data quality issues can lead to either downtime or wasted productivity. We believe that executives will search for automated solutions to observe, detect, and enforce data quality standards across services at the source, before it goes to downstream systems.
Ustin Zarubin, CEO & Co-Founder of Streamdal
On Single-Box Apps: The single-box app, ushered in by the container revolution, will begin to go the way of the single-page app. Overfixation on keeping everything inside single boxes will give way to the coordination, automation, and developer experience opportunities when this assumption is relaxed. All the single-box apps aren't going anywhere, but infrastructure which is too entrenched in single-box will lose steam, with growing adoption of abstractions and platforms which can natively handle single box, multi-box, or no box within a single broader program. Temporal, Ray, and Dapr are all strong, but early indications of this momentum.
Donny Greenberg, CEO & Co-Founder of Runhouse
On Security Automation: The velocity of progress as the industry moves from DevOps into CI/CD and serverless continues at a frantic pace. The adoption of fully automated security tooling into the developer-driven process is the natural step here. By 2026, companies that have not adopted AI-driven security automation solutions for both delivery of quality product and timeliness to the marketplace will be at a significant -- and likely existential -- competitive disadvantage.
James Wickett, CEO & Co-Founder of DryRun Security
If you’re an early-stage enterprise founder or operator — connect with us directly to chat about anything GTM or check out our events page to stay in the loop on all things happening in the Work-Bench community.
