How AI is changing the on-call experience for startups

When innovation meets empathy, magic happens. I came across the story of Priyanka Kulkarni, a former Microsoft machine learning scientist, and it honestly made me pause. After spending years navigating the complex U.S. visa process herself, she decided to solve it - not by adding more paperwork or consultants, but by building tech that simplifies the entire journey. That’s how Casium was born - an AI-powered startup that helps professionals and employers complete employment-based visa processes in under 10 days (something that used to take 3–6 months!). What impressed me most is how Casium automates everything - from assessments to compliance and filings - removing human errors, delays, and confusion that so many global talents face. Priyanka’s background is just as inspiring with degrees from the University of Mumbai and Johns Hopkins University, and nearly a decade at Microsoft, shaping AI strategy for products like Office. She’s now using all that experience to fix one of the most frustrating systems out there immigration. Today, Casium, backed by the AI2 Incubator’s 2024 cohort in Seattle, is helping hundreds of candidates with visa assessments, filings, and compliance reviews boasting an exceptional approval rate. It’s incredible to see technology being used not just for convenience, but for fairness, clarity, and opportunity. More power to founders who build solutions from lived experiences because that’s where the most meaningful innovation begin.

AI in the Immigration Space: Casium's Approach to Work Visa Process Priyanka Kulkarni, a former Microsoft machine learning scientist, has founded Casium, a startup aimed at simplifying the U.S. employment-based visa process through AI. Drawing from her own experience with the H-1B visa system, Priyanka developed a platform that automates visa assessments, compliance reviews, and filings, significantly reducing processing times from several months to under 10 business days. This solution looks to streamline what has traditionally been a complex, manual process. Casium recently raised $5M in seed funding to further develop their platform and continue improving the visa filing process for companies. You can read more about Casium’s innovative approach here: https://lnkd.in/gHWSxbrB #AI #Startup #VisaTech #MachineLearning #Innovation #Casium #TechForGood

AI in the Immigration Space: Casium's Approach to Work Visa Process Priyanka Kulkarni, a former Microsoft machine learning scientist, has founded Casium, a startup aimed at simplifying the U.S. employment-based visa process through AI. Drawing from her own experience with the H-1B visa system, Priyanka developed a platform that automates visa assessments, compliance reviews, and filings, significantly reducing processing times from several months to under 10 business days. This solution looks to streamline what has traditionally been a complex, manual process. Casium recently raised $5M in seed funding to further develop their platform and continue improving the visa filing process for companies. You can read more about Casium’s innovative approach here: https://lnkd.in/gNejbVsd #AI #Startup #VisaTech #MachineLearning #Innovation #Casium #TechForGood

About 4 months back, we were in the final stage of negotiation with a client. We had understood the requirements, shared the proposal, and were almost there. Until one of the decision makers jumped on a call and said, “Can you do the same thing using Lovable for a much lower rate? Otherwise, we’ll just build it ourselves using this.” We politely declined. As far as I know, that project still isn’t live. I’m not against AI tools or no-code platforms, they’re incredible for what they’re meant to do. They can’t replace someone who understands edge cases, scalability, or user behaviour beyond what a prompt can predict. Tools can accelerate good ideas. But only great engineers turn them into great products. #TechTalent #Engineers #ProblemSolvers #AiRecruiting Hire problem solvers and not just coders - RemoteEngine

Did AI truly take your job, or did your employer just use AI as an excuse? Over the past year, thousands of tech professionals have been informed that their roles have been eliminated due to “AI automation.” But here’s what’s really happening behind the curtain: Many of these companies are quietly rehiring employees, often for the same roles, through H-1B visa programs and third-party staffing agencies. AI didn’t replace all those jobs. Cost-cutting and clever storytelling did. As leaders, engineers, and innovators, we need to separate genuine automation progress from corporate PR spin. Because when we let “AI” take the blame, we overlook the fundamental issue of how talent, both domestic and foreign, is being undervalued and misrepresented. Transparency matters. Ethics matter. And it’s time we started holding companies accountable for how they use both AI and H-1B hiring in the name of “innovation.” https://lnkd.in/gVTH_FaZ #AIandWorkforce #TechLayoffs #H1BVisa #FutureOfWork #CorporateEthics #AI #ConfigrTechnologies

Did AI truly take your job, or did your employer just use AI as an excuse? Over the past year, thousands of tech professionals have been informed that their roles have been eliminated due to “AI automation.” But here’s what’s really happening behind the curtain: Many of these companies are quietly rehiring employees, often for the same roles, through H-1B visa programs and third-party staffing agencies. AI didn’t replace all those jobs. Cost-cutting and clever storytelling did. As leaders, engineers, and innovators, we need to separate genuine automation progress from corporate PR spin. Because when we let “AI” take the blame, we overlook the fundamental issue of how talent, both domestic and foreign, is being undervalued and misrepresented. Transparency matters. Ethics matter. And it’s time we started holding companies accountable for how they use both AI and H-1B hiring in the name of “innovation.” https://lnkd.in/gGBj7Vh6 #AIandWorkforce #TechLayoffs #H1BVisa #FutureOfWork #CorporateEthics #AI #ConfigrTechnologies

After nearly a decade at Microsoft as a machine learning scientist, Priyanka Kulkarni has founded Casium, an AI-driven platform transforming how organizations manage employment-based immigration. Designed to simplify one of the most complex administrative processes, Casium replaces outdated, manual visa management systems with intelligent automation. The platform enables employers to handle visa cases end-to-end, cutting processing timelines from 3–6 months to under 10 business days. By integrating AI-powered document verification, compliance tracking, and workflow management, Casium ensures accuracy, transparency, and faster decision-making. Since launch, it has already assisted hundreds of professionals through assessments, compliance reviews, and filings, achieving an impressive approval rate. Casium offers free initial assessments and a flat-fee filing model, with a subscription service underway. Born and raised in India, Priyanka holds degrees in Computer Engineering (University of Mumbai) and Applied Mathematics (Johns Hopkins University). Having personally navigated the complexities of the U.S. visa system, she built Casium to empower skilled professionals with speed, clarity, and confidence. By merging AI and immigration expertise, Casium is setting a new benchmark for digital transformation in global mobility. #Casium #PriyankaKulkarni #ImmigrationTech #AIInnovation #DigitalTransformation #VisaAutomation #LegalTech #GlobalTalent #AIStartup #WorkforceMobility

The Evolution of AI External Calls: Tools → MCP → Skills Claude has just released a new feature - skills. It looks like a stable trend to evolve external calls from AI clients. We are driving to the world where consumers just "get" solutions in communication with their preferred AI and software businesses just provide federated solutions which can give your AI client puzzle pieces helping crack it quickly and make your happy. We started from tools for agents which were available only for those who can create a good system prompt and handle the response correctly. It required some programming not only for tools by itself but also for orchestration. Your code was in charge of agent's decision handling. Now we have a lot of programming/no code frameworks. Which simplify that process. Then MCP (Model Context Protocol) has appeared as the answer for not standardised tool calls. You as an enthusiast-programmer can create MCP server and your user's friendly MCP client will call it. Experience of connection is still a bit awkward. The best option we've got on this level -> access to MCP via URL but it feels a bit too technical. The worst experience -> access local MCP (likely installed via npm). Now - skills. You still can install them either locally or connect available remote ones but now they are more about your apps like sheets or slides. We will see skills from Booking, skills from Uber, skills from every business which want to stay relevant in this world of AI-gated access to the Internet. Obviously, a marketplace of skills. All this looks like strong trend to future the Internet of Skills. Anthropic's post -> https://lnkd.in/dnGgCYKx for founders and business owners -> https://lnkd.in/dVP6StWB

AI is fragmenting workforce pipelines. Here’s an example, it’s increasing the need for senior software developers while reducing the need for entry level software developers. This has long term impacts on career pipelines. Historically, technology advancement creates more new jobs. But what happens when those jobs aren’t attainable?

The Entry-Level Tech Job Is Officially Dead If we replace all the junior tech workers, we’re going to run out of experienced tech workers #entrepreneurship #technology #jobs #money #business You can make the argument that AI eliminating jobs is actually a good thing. You can make that argument, sure, but that argument is wrong. I know. I’ve argued it. From both sides. About six months ago, I wrote a post that claimed to be a warning as to which types of jobs that AI was going to replace, but it was really a warning that the “type” of job AI was coming for didn’t matter. It was becoming obvious that Corporate Tech, in its frantic rush to profitability, was doing the thing with AI that it does with every new technology. It exploits the new tech as a chance to finally get rid of all those annoying tech workers – with their cavalier fashion sense and their video game references no one understands. Well, it finally happened. We got rid of all the junior tech workers. Here’s what will happen next. Read on... https://lnkd.in/em-X-DpN