What Does an Agentic SDLC Look Like in a Real Company?

​Everyone's asking what engineering looks like once you move from operating agents to a fully agentic SDLC. In this session you'll learn from a team already working that way.
​Register to attend. We also train R&D teams to build Agentic SDLC in-house - see what the course covers.
​The R&D team of Ironscales, an AI-driven, self-learning email security platform, built an agentic SDLC that runs in production today.

It's engineered as a chain of specialized agents handling requirements, architecture, code, test environments, and QA, with a human sign-off gating every step before work moves forward. More than 90 workflows have moved through it, real features and real PRs.

Join us to learn the thinking, methods, and challenges behind building an AI SDLC workflow, and what it changed about how a company's teams work together, so you can start planning how to build one at your own company.

The session will be hosted by Jonathan Yarkoni, CEO & Co-founder of Latent AI, with guests Dima Yedunov, AI Operations Manager, and Vasili Anoshin, Backend Engineer, at Ironscales.
​Language: English
​🚀 What we'll cover

• ​The eight stages of a working SDLC, where the human gates sit, and what it takes to clear one
• ​Grill-me and circle-back: how the pipeline catches a gap in an earlier stage and folds back on itself without spinning forever
• ​Why a deterministic state machine, not the agents, decides what runs next
• ​How QA agents reason over a live map of the product instead of re-running everything
• ​What it takes to grow and scale a system like this

​**✨ You should attend if:**

• ​Your team ships with AI assistance, but everything past the first draft is still manual
• ​You've built an agent pipeline that works in a demo and stalls before production
• ​You're deciding whether to bet on an agentic lifecycle and want the real cost from someone running one

​This session is the teardown. If you want your own team building one, our AI SDLC course covers it end to end - here's the full syllabus.

​FAQs

• ​Is this technical? Yes. The session goes into the actual mechanism: state machines, approval gates, QA loops.
• ​Will it be recorded? The session is live and unrecorded, so the Q&A stays candid. Registered attendees receive a takeaways summary afterward.
• ​Is this a product pitch for Ironscales? No. It's a working teardown of a system they run internally.
• ​I can't attend. Can you let me know about future events? Definitely. Register anyway and we'll send you the takeaways plus a heads-up on the next one.
• ​Can you train my team to build this? Yes - that's our AI SDLC course. Here's what it covers.

​About Latent AI
​Latent AI is a senior AI team that designs, builds, and deploys production-grade AI systems for enterprises. Drawing on global talent, we deliver measurable outcomes in real-world environments. With clients across sectors, including Discount Bank, Yad2, Grow, and Lightico, we build and deploy AI systems that perform reliably at scale.
​Our Academia arm carries that experience into training, equipping teams with the tools for real AI transformation. It has trained dozens of technology companies globally, including Appsflyer, Palo Alto Networks, Mobileye and more.
​Looking forward to having you join us.