Lantern

Humanizing dementia care with AI. Through Lantern's Do Good program, we partnered with Optima Living to transform the memory care intake experience with the AI-powered Spark App. The result? ◾ AI-generated Living Stories that help personalize care from day one ◾ Reduced administrative burden so caregivers can spend more time with residents and families ◾ A new model for person-centered dementia care Learn how AI and compassion came together to transform the memory care experience: https://loom.ly/nr5qtF8

If your Copilot Studio knowledge agent is struggling to return strong answers, the issue may not be the prompt. A lot of agent performance comes down to retrieval: how the system finds, ranks, and surfaces information before it ever generates a response. That becomes especially important when teams are working with long, complex documents where the right answer may be buried below the surface. This carousel breaks down why that happens, what teams often misunderstand, and why retrieval depth deserves more attention in agent design. Read the full blog post: https://lnkd.in/ge4Zjssz #AI #CopilotStudio #MicrosoftCopilot #GenerativeAI #EnterpriseAI

GitHub’s pricing shift introduces a new operational reality for Copilot users. Credits now pool across the organization, agentic features consume those credits differently, and exhausted pools can lead to blocked usage or incremental spend. In this Lantern Minute, Principal GitHub AI Solutions Consultant Eric Smith explains why governance matters more under the new model.

Lantern is proud to welcome our newest team member, Yakov Yuzhakov! Yakov brings over 14 years of professional experience in data engineering, analytics, and software solution development, with extensive expertise in designing and building enterprise data platforms. Outside of work, Yakov loves spending quality time with his family and his two daughters.

The first wave of AI coding made it easy to generate quick prototypes. But enterprise software development asks for more than a good prompt. It requires planning, testing, iteration, security, governance, and the ability to work inside real engineering workflows. That’s the shift this carousel explores: why vibe-driven coding hits a ceiling, what agentic development actually means, and how the developer role changes alongside it. If you’re thinking about AI in software engineering beyond one-off code generation, the full blog is worth a read. #GitHubCopilot #AgenticAI #SoftwareEngineering #EnterpriseAI #AIDevelopment

Bugs caught in production cost roughly 100x more to fix than the same bug caught at design. And on average, about 15% of defects still slip through QA and reach it. That gap isn't a testing effort problem. It's a testing model problem. The usual fix, more test automation, comes with its own tax: scripts only developers can maintain, authentication flows that are painful to automate, and suites that become their own maintenance project as the app changes. Teams end up with partial coverage and a test suite that feels like a liability. We think the answer isn't more automation. It's a better split between what AI should own and what still needs a human. In this blog post, we break down how Lantern does this with tools most teams already have (VS Code, GitHub Copilot, Playwright), starting from something every team already writes: acceptance criteria. No new platform. No scripting expertise required. No humans removed from the decisions that need one. See what QA looks like when AI handles the repetition and your team handles the judgment. Read the full post: https://loom.ly/LGTheik

Most organizations don't struggle to find technology partners. They struggle to find a partner that works the way they need them to. ✅Some teams know exactly what needs to be done and simply need the right expertise. ✅Others know where they want to go but expect priorities to evolve as they learn. ✅ And some want a trusted partner to take ownership and deliver a defined outcome. That's why we've built three engagement models at Lantern: 1️⃣ Lantern Squads - We work together to drive outcomes. A dedicated cross-functional team focused on progress, learning, and adapting as priorities change. Instead of locking everything down upfront, we continuously prioritize work based on the outcomes that matter most. "We know where we're heading, but we need a partner who can adapt with us along the way." 2️⃣ Lantern Led - We own delivery and are accountable for outcomes. When you have a defined initiative and want a single partner responsible for planning, execution, and delivery from end to end. "We have a clear objective. We want a trusted partner to take it from here." 3️⃣ Lantern Talent - You lead the work. We provide the expertise. Ideal when you have a clear roadmap and need specialized skills, additional capacity, or experienced practitioners embedded within your team. "We know what needs to be done. We need the right people to help us do it." ------------ The right engagement model isn't determined by company size, budget, or industry. It's determined by a simple question: How much ownership do you want to retain, and how much do you want your partner to take on? Because the reality is that business needs change. The best engagement model today may not be the right one six months from now. That's why flexibility isn't a feature of how we work, it's the foundation.

Lantern is proud to welcome our newest team member, Mokshita Madan! Mokshita brings extensive experience in leveraging data to drive business value. She has helped organizations make informed decisions, optimize operations, and uncover growth opportunities, while effectively bridging business needs with technical execution. Outside of work, she enjoys yoga, nature walks, beach visits, and occasional deep ‘life discussions’ with her very logical 5-year-old Golden Retriever, while leveraging her curiosity in AI to inspire innovation and continuous improvement for personal and professional growth.

We are doing it again! We are launching our Lantern Early Career Program, built to help curious people accelerate learning through mentorship, hands-on experience, and exposure to modern technologies. The world of technology has changed drastically over the past couple of years, with no slowdown in sight. This is a great way to immerse yourself right in the middle of the action! Come join our teams, leading the way on all things AI across Microsoft and GitHub. Whether you're launching your career or figuring out what comes next, growth rarely happens by staying comfortable. Let's grow together! Applications are open.  Dallas: https://lnkd.in/ePi_wPMK  Edmonton: https://lnkd.in/eWnUtaaC 

AI cost risk rarely shows up in the demo environment. It becomes visible when users begin engaging across the business and organizations are forced to manage quotas, policies, and operational ownership at scale. In this Lantern Minute, Practice Lead, AI Foundations, Jason Davis discusses why pilots can create false confidence around cost and what needs to be in place before production.