Mukund Balasubramanian

We’re excited to partner with Cognition, the stellar crew behind Devin 🤖 and Windsurf 🌊, to make SWE agents more effective. Here’s the workflow: ✨ CoreStory extracts the hidden business logic, requirements, and complex behaviors in legacy codebases into a persistent spec. ⚡ You pass that spec as context and tasks to Devin, Cognition's Autonomous AI Software Engineer. 🚀 With the spec to navigate, the AI agent gets more powerful. It can modernize entire systems, or use the spec to field even more tickets with higher confidence. This is what the future of software engineering looks like: Human devs + AI agents, working side by side, with CoreStory fueling the spec. #AI #Agents #SpecDrivenDevelopment #LegacyModernization #CoreStory #Cognition

17

Excited to be speaking at Innovate QA 2026 this June in Seattle. Over the past two decades across companies like Microsoft, Amazon, Google, and now building FlintLab, I’ve seen how testing has evolved from manual validation to automation frameworks, and now toward AI-driven systems. In my session, “Is Today’s Test Infra Ready for AI? The Hard Questions No One Is Asking,” I’ll share perspectives on why AI-driven testing requires a fundamentally different infrastructure mindset and what engineering teams should rethink before scaling AI agents. Looking forward to connecting with fellow engineers, quality leaders, and practitioners pushing the boundaries of modern test infrastructure. See you in Seattle on June 4–5. #InnovateQA #QualityEngineering #AITesting #TestInfrastructure #SoftwareQuality

31

Vianai Systems, Inc.'s badass CTO, Navin Budhiraja, doing his thing. Insightful interview on the technological, professional and societal implications of AI and the need for enterprise grade AI technology to truly drive business performance.

2

#AIJourneyWithMaheshRaja post #8 its been a while I wrote on my core ( AI ) Most of you know that like in any organization we at Conscendo paving our way to become an AI driven organization on par with Industrial paradigm shift that is happening, We too feel the pressure of "AI FOMO ( Feelingof Missing Out )." We see the headlines and feel we need to implement something immediately. But rushing in without a roadmap often leads to failed POCs and, more importantly, lot of invalid learning and may be a risk-scoring of breach of customer trust Here are my three main takeaways on how to approach AI the right way: 1. Identify Bottlenecks, 2. Plan for Fool proof Trust Layer 3. Nominate an AI focal point in the organization 1) In my experience with AI, The most successful AI implementations don’t start with the technology; they start with the team on the ground, check "Where and What are our bottlenecks", 2.) If AI team can’t guarantee that our proprietary data won’t be used to train public models, the risk to our brand is too high. 3) Appoint an AI Champion AI isn't just an "IT project." It requires a Single Point of Contact (SPOC) who understands both the business goals and the technical limitations. Without a dedicated champion to navigate the "hallucinations" and refine the outputs, projects often stall at the demo stage. Let’s stop racing to be first and start racing to be the most reliable. I’m curious to hear from my network-are you currently testing AI in your workflows? What has been your biggest hurdle so far? #ArtificialIntelligence #SMB #DigitalTransformation #BusinessStrategy #Leadership #AIAdaption #Salesforce #ConscendoTechnologies

14

1 Comment

As engineering leaders, we’re constantly balancing speed and reliability. This Just-in-Time testing approach solves a burgeoning, critical efficiency problem that all high velocity teams are facing or will face imminently. As we demonstrated in our last paper, traditional automated test generation – the latest, greatest software development innovation – focuses on “hardening,” or creating tests that pass today to protect against future regressions. At Meta, Mark Harman and his team have shifted the paradigm toward “catching,” where generated tests are designed to fail on proposed changes, surfacing bugs before they ever land in the codebase. Why this matters: * Zero maintenance overhead → no more brittle test suites * Scales with rapid development → adapts instantly to code changes * Focuses engineering time where it matters most – on real bugs, not false positives * Proven results → production issues prevented with minimal drag This shift from “hardening” tests to intelligent “catching” tests is exactly the kind of innovation we need to keep pace with agentic development. Looking forward to seeing how this transforms testing practices across the industry. Check out Mark and team’s blog and paper to learn more! https://lnkd.in/gUY4JddR

62

2 Comments