Agent-driven workflows were key to engineering unlocking the value of AI and mov
By Observe Inc · October 27, 2026 · Curated by George's Blog
Agent-driven workflows were key to engineering unlocking the value of AI and moving faster at 6sense.
Moving from engineers using AI-assisted tools in their SDLC to their AI-native "AI-SDLC" agent was the step that increased productivity and sped up delivery, as Viral Bajaria, Chief Innovation Officer, shared on Snowflake's The CTO Circle Podcast.
The AI SDLC agents use a living shared document that provides context and an immutable decision log. They move work through the software development process, calling humans into the loop as needed, filing tickets, and tagging folks in Slack. Humans no longer need to log in to Jira, GitHub, and other tools; the context comes with them.
AI SDLC enables the 6sense team to go from vision to customer-ready prototype in 2.5 weeks, with a prototype that's already integrated into their platform services. Viral joins Shruti Bhat and John Mark Troyer on our podcast to explain how it works — a dynamic team structure, aggregated context across tools, and an agent harness that adds observability to track decisions and costs per task. The team can measure their productivity gains in a concrete way that wasn't apparent at the start, when they began using AI-assisted coding tools like Cursor and Claude Code.
The real leverage came when they added these AI-native abstractions at the SDLC process level, not when they added AI-native tools at the individual level using the same SDLC process.
Check out the full episode of Inside the CTO Circle 👉 https://lnkd.in/ehHExMZt
#aiengineering #engineeringleadership #SDLC #agenticworkflow