Andy Ratsirason and Shalini Aggarwal, both former Amazon employees, left the tech giant at different times and eventually reunited to co-found Tenafli, an AI startup focused on creating intelligent companions for older adults. Their journey highlights the mental and operational shifts required when transitioning from a large corporation to the unpredictable world of early-stage startups. Shalini, a 50-year-old CEO based in San Jose, California, joined Amazon in 2015 after relocating from India. Andy, a 37-year-old CTO, first joined Amazon in 2014, drawn by the Silicon Valley ecosystem and his entrepreneurial ambitions. He left briefly, returned in 2020, and departed again in 2023. After his second exit, he began exploring startup opportunities and noticed Shalini’s active presence on LinkedIn—posting about risk-taking and entrepreneurship. He reached out, and she soon joined him as cofounder. Their biggest realization early on was how much they had taken for granted at Amazon. In enterprise environments, infrastructure, processes, and teams are already in place. In a startup, everything must be built from scratch. They had to unlearn the Big Tech mindset of building first and asking questions later. Shalini’s decision to leave Amazon was deeply personal. After her mother passed during the pandemic and her father became isolated, she began thinking about how technology could help seniors stay engaged. Her work on AI-driven music recommendations inspired her to create a product that could offer personalized activity suggestions and companionship. The AI boom, combined with her approaching 50th birthday, made the timing feel right. In September 2024, she submitted her resignation. Andy, meanwhile, had been weighing a similar choice. After nearly three years back at Amazon, he chose to leave in February 2023 to pursue his own venture. Their shared vision and complementary skills made the partnership a natural fit. One of the most difficult shifts was moving from a “build-first” mentality to a “validate-first” approach. After months of development, their initial product launch saw little traction. The wake-up call came when almost no users showed up organically. They learned that shipping a product isn’t the goal—proving demand is. They began conducting customer interviews early, testing interest through waitlists, community posts, and pilot partnerships before investing heavily in development. This customer-centric approach became central to their process. Frugality became another key lesson. Despite receiving AWS credits and other startup support, they quickly overspent by over-provisioning resources and leaving systems running. They implemented budget alerts, automated shutdowns, and began using a local machine for testing. This hands-on cost management taught them to build lean and scale only when necessary. AI also became a game-changer. Instead of spending hours reading research, they use AI tools to scan content, extract key insights, and summarize what matters. This freed up time to attend industry events, engage with users, and refine their product based on real feedback. AI also reduced the need for certain roles. It acts like a junior engineer, handling coding tasks based on clear requirements. Shalini no longer needs a dedicated UI designer—she uses AI to prototype and iterate quickly. Andy notes that what once required a large team and significant funding can now be achieved with just two engineers and cloud subscriptions. Their Big Tech backgrounds helped them structure processes and avoid chaos, but also held them back. They had to overcome the assumption that infrastructure and tools would always be available. They learned to prioritize learning and validation over polish. The hardest lesson was overcoming the fear of launching something imperfect. At Amazon, perfection was expected. In startups, the real risk isn’t rough edges—it’s building something nobody wants. Their journey underscores a critical truth: success in startups isn’t about replicating corporate habits. It’s about adapting, listening, and iterating faster than ever before.