Perplexity AI Business Fellowship Wrapped

Connected Intelligence : The Feedback Loop Podcast

0:00

-40:15

Perplexity AI Business Fellowship Wrapped

The Fellowship Reflections and Future Directions as discussed in August 9th 2025 90 minutes wrap-up call. here's the audio / video recaps and key points of focus made.

Poonam Parihar

Aug 13, 2025

Transcript

The "Perplexity Fellowship Wrap-up: Looking Back, Moving Forward" call, the final one in the a monthly series of Online catch-up calls during the 6 months of fellowship, was attended by Adarsh S Lathika, Ángel Molina, Chon “Choon” Chua, Daniela Busse, Ph.D., DELIA GREEN, Giulia Felicetti, John Fawole, Khusshboo Mehta, nishant gupta, Ravi Vinnakota, shakthi prashanth, Siddharth Nandagopal, Srinivasamurthy Mohan, Stephanie Goutos, Sukhdeep Singh, Tech learning, Uvika Sharma, and Yogesh Panchal and hosted by Poonam Parihar.

The Audio podcast above provides a great recap of the conversation as it flew from reflections to serious discussion on the future direction of the AI and participants alike.

Core Quotes and Insights from the Fellows

Here are the key quotes/points made by each participant:

Poonam Parihar – Driving Ongoing Community Conversations

Poonam Parihar organized and facilitated the recurring calls, viewing them as fun, free‑flowing chats where AI topics naturally emerged. She said the Perplexity Fellowship had generated a lot of “PR pumping” for the company, effectively making fellows “guinea pigs” for testing new features, “pre‑customer research” and product testing. She advocated continuing community conversations, such as this concluding series and starting a new focussed series with a brand new perspective and format, to keep up with the rapid pace of AI evolution, which she expected would stabilize by 2026 or 2027. She also shared that she was working on a deep‑tech project involving AI 5G implementation with AI chips (NPUs), which she viewed as a game‑changer for on‑device processing.

Uvika Sharma – Learning 10x and Advocating for Women in AI

Uvika Sharma, an “AI tool junkie” but cautious about data privacy, had already been a Perplexity Pro user before the fellowship. She described the fellowship journey as “phenomenal” and said her learning had increased “10x” through interactions and diverse perspectives. She valued hearing “uninterrupted wisdom” from leaders like Jensen Huang and Ali and expressed a desire to see more women in tech and AI—specifically a female AI leader—in fireside chats. She agreed with Poonam and found the fellowship model “very clever”. She also expressed discomfort and concern about connecting Comet to her Gmail due to privacy implications, especially regarding private contracts.

Chon “Choon” Chua – Inspired by Leaders and Diversity

Chon “Choon” Chua expressed gratitude for the fellowship’s “amazing exposure” and “lineup of the speakers,” noting that the organizers had likely leveraged their social networks to attract top leaders like Jensen Huang. She was impressed by Huang’s “unfiltered” and “heartwarming” character. She mentioned that she had switched from using Anthropic Claude as her main tool to using Perplexity “a lot more” and that she “really loved the concept of Comet.” She also appreciated the diversity of talent within the fellowship.

Stephanie Goutos – Balancing AI Benefits with Privacy Risks

Stephanie Goutos, an attorney and limited partner in a venture fund, shared that she had connected Comet to her personal Gmail, accepting the privacy risk because the time‑saving benefits outweighed her concerns, but she had not connected it to her work email. She agreed that the fellowship was “really smart” for Perplexity, generating “a ton of free publicity” and exposure across industries.

Ravi Vinnakota – Championing Single‑Person Startups

Ravi Vinnakota, initially suspicious of the fellowship’s value, said he had ultimately found the journey very good. As an engineering head, he had introduced AI tools like ChatGPT from day one for product requirements, code generation, and deployments. Inspired by Roy Bahat’s session, he was passionate about building “single‑person billion‑dollar startups” or “as lean as possible” teams, believing AI made this feasible. He predicted AI tools would stabilize by 2026 or 2027, similar to how organizations settle on cloud providers, and he strongly believed “agentic AI” would dominate by then, performing at least half of human work.

Khusshboo Mehta – From Prompt Engineering to AI Literacy

Khusshboo Mehta described the fellowship as a “phenomenal journey” and noted the unusually active WhatsApp group. She highlighted the shift from prompt engineering to context engineering and the access to tools like Comet. She emphasized that “AI literacy training” was crucial groundwork for AI adoption within teams and organizations.

Adarsh S Lathika – Privacy‑Focused Builder of Personal AI Apps

Adarsh S Lathika said he actively practiced digital privacy by using a secondary phone for social media and avoiding financial apps on his phone. He seconded Poonam on building personal AI apps and informed he’s been working on for his own needs, such as an apartment management system and games he used to play. He found Perplexity did a “much better job in report generation” than ChatGPT, though it occasionally hallucinated, requiring data validation. He stressed that AI output quality depended on the user’s input quality, domain expertise, and effective prompt design.

Yogesh Panchal – Balancing Data Sharing and Productivity

Yogesh Panchal agreed with the view that Perplexity’s PR strategy had been successful. He said he was willing to share personal data with AI tools if it increased productivity and improved the technology, noting that “technology always asks something from us.” He supported individuals building personal AI apps for their own use, likening it to customizing one’s own car. He also suggested using a “burner” email address for privacy‑sensitive AI interactions.

Delia Green – Humility and Hands‑On AI Adoption

Delia Green, coming from a finance and accounting background, said she had developed a growing passion for technology. She found the fellowship gave her exposure to numerous AI tools she had not known about and greatly appreciated the community’s support in solving specific issues, such as automating emails using Make.com. She learned the importance of humility and keeping an open mind (“keep down your ego”) since everyone could learn something new. She planned to apply AI in her daily corporate tasks and was also considering opening her own venture. She noted being surprised by the very fast pace of AI change.

Sukhdeep Singh – Bringing AI to Traditional Industries

Sukhdeep Singh, a sales professional from the consumer goods industry, described the fellowship as a “zero‑to‑one” learning journey, calling it “100x phenomenal.” Before the fellowship, he had not heard of Perplexity and had only used basic ChatGPT. He now actively bookmarked and made efforts to apply all mentioned AI tools to his work. He agreed with Poonam that there was huge potential for AI beyond the tech sector, especially in “primitive industries” like mining, natural resources, and agriculture. He saw AI combined with IoT as a “killer” combination in the physical world and had started learning about IoT. He regarded AI chips (NPUs) as a game‑changer for hardware integration.

Srinivasamurthy Mohan – Learning Through Weekly Peer Calls

Srinivasamurthy Mohan, from a non‑technical background in sales operations and pricing, did not work with programming languages but understood the tech landscape. He had been exploring multiple AI tools and new topics like Copilot Studio and Make.com to learn. He emphasized the value of weekly calls for “learning by hearing from others,” which, he said, helped condense information that would otherwise take hours to find.

Giulia Felicetti – Bringing Perplexity to the Italian Market

Giulia Felicetti had been directly involved in bringing Perplexity to the mass market in Italy and had witnessed the product’s significant growth. She was impressed by the quality of people on both the Perplexity side and among the fellows. She admitted finding it challenging to stay updated in AI but recommended habits like listening to podcasts and, most importantly, actively using the tools. She also highlighted the value of taking notes and bookmarking important information from the WhatsApp group.

10 Reflection Points

Here are 10 unique reflection points based on the live discussion, offering consolidated insights and observations from participants:

Transformative Fellowship Learning Journey
The Perplexity Fellowship was described as "phenomenal" and a profoundly transformative experience that accelerated learning by "10x"–"100x". Exposure to leaders like Jensen Huang and Ali provided "uninterrupted wisdom" that inspired participants to explore AI more deeply.
Active and Supportive Community
The highly engaged WhatsApp group became a cornerstone of the fellowship’s success—fostering friendships, trust, and collaboration while offering diverse perspectives and mutual help.
Perplexity’s Strategic Fellowship Design
Participants saw the fellowship as a clever strategic move—doubling as “pre-customer research” and a live product-testing ground for features like Labs and Comet, generating valuable feedback and significant free publicity.
Managing the Rapid Pace of AI Evolution
With new tools and models emerging almost weekly, participants found it challenging to keep up and differentiate genuine advances from hype cycles. Many expect the market to enter a stabilization phase by 2026–2027.
Rise of Agentic AI and Lean Innovation
By 2026, agentic AI is predicted to perform at least half of human work, enabling “single-person startups” and widespread development of personal AI apps tailored to specific needs such as smart home systems or niche applications.
AI’s Expansion into Traditional and ‘Primitive’ Industries
Beyond the tech sector, AI is expected to revolutionize industries like agriculture, mining, and natural resources, enabling applications from precision farming to resource optimization.
Convergence of AI with Hardware
AI chips (NPUs) combined with IoT and 5G are viewed as a game-changer for on-device processing, unlocking faster, more private, and location-aware AI applications in physical environments.
From Flashy Demos to Daily Productivity
There is a need to shift AI narrative and marketing from one-off novelty use cases to demonstrating how AI can deliver tangible time savings and productivity boosts in daily professional work.
Privacy Concerns and Adoption
Attitudes toward connecting AI tools to sensitive accounts like Gmail vary. Some prioritize convenience, others use measures like burner accounts, and many stress the need for stronger privacy safeguards.
AI Literacy as a Foundational Skill
Equipping people and teams with foundational AI knowledge is seen as critical groundwork for successful adoption. AI literacy enables confident use, informed decision-making, and leadership in AI integration.

10 Key Future Directions & Learnings.

Participants identified the following forward-looking themes for AI’s evolution, adoption, and continued learning:

Stabilization of AI Tool Ecosystem – Moving from rapid churn toward a more stable set of trusted tools by the mid-2020s.
Dominance of Agentic AI – Significant automation of human workflows, leading to marketplaces for independent AI agents.
Cross-Industry AI Integration – Deep adoption across both high-tech and traditional sectors.
Hardware-Enhanced AI Capabilities – NPUs, 5G, and IoT driving embedded, real-world AI uses.
Focus on Personal AI Applications – Individuals creating AI-driven tools for their unique needs.
Marketing AI for Everyday Workflows – Showing practical, productivity-driven results rather than just novel demos.
Evolving Privacy Standards & Trust Models – Balancing convenience with data protection.
Organization-Wide AI Literacy Initiatives – Structured training to embed AI readiness.
Community-Driven Learning Networks – Sustaining the collaboration and knowledge-sharing seen in the fellowship.
Prioritizing Real-World Value Over Hype – Critical focus on proven impact and measurable benefits.

What fellows are doing to keep up with the evolution of AI and their take ways for both personal and professional development in AI that you could use

Individual Commitment to Continuous AI Learning and Application: Participants are actively planning to keep learning and applying new AI tools and models to their work and personal lives. This involves dedicating time to test and use the tools as they emerge. Delia Green, for example, is keen to understand how to apply AI in her daily corporate tasks and potentially open her own venture. Sukhdeep Singh, a sales professional, is making a conscious effort to bookmark and use all mentioned tools to apply them to his work, recognizing AI's potential beyond tech. Srinivasamurthy Mohan, from a non-technical background, plans to continue exploring multiple AI tools and new topics like Copilot Studio and Make.com to gain learnings.
Focus on Building Personal AI Applications and Lean Startups: A significant future direction is the idea of individuals building "personal AI apps" for their own needs, rather than solely for commercial purposes. This includes managing personal aspects like apartment systems or creating custom games. This aligns with the concept of "single-person startups" or "as lean as possible" teams, empowered by AI agents, as championed by Ravi Vinnakota, who is passionate about this model.
Strategic AI Literacy Training for Organizations: For broader AI adoption, a crucial future step involves establishing "AI literacy training" as foundational groundwork within organizations. This indicates a recognition of the need to equip teams with a basic understanding of AI to facilitate its successful integration and "AI leadership" across various contexts.
·Anticipating the Stabilization and Dominance of Agentic AI: Looking ahead, participants anticipate a stabilization phase in the AI tools market by 2026 or 2027, where organizations might settle on specific tools rather than constantly switching. More significantly, agentic AI is expected to heavily dominate the world, with real-world use cases emerging by 2026, potentially performing "at least 50% of human work". This suggests a future where marketplaces for independent agents could become prevalent.
Expansion of AI into Diverse Industries and Hardware Integration: The future plans for AI involve its expansion beyond the traditional tech sector, significantly impacting "primitive industries" like mining, natural resources, and agriculture. This widespread integration includes the convergence of AI capabilities with hardware, particularly AI chips (NPUs), which is seen as a "game-changer" for on-device processing and physical world applications, especially when combined with 5G and IoT. Poonam Parihar mentioned working on a project involving AI 5G implementation with AI chips.
Shifting Focus in AI Marketing to Daily Productivity: There is a strong call for AI development and marketing to shift from showcasing "one-time, fancy" use cases (like planning a wedding or remodeling a kitchen) to demonstrating how AI can tangibly improve "daily professional tasks" and boost productivity, saving hours in an "8 to 10 hours of working day". This reflects a desire for AI to become a practical tool for efficient daily operations in both personal and business contexts.
Continuation of Community-Driven Learning and Interaction: To keep up with the rapid pace of AI development, there is a clear plan and desire to continue the community calls on a regular basis. This rhythm is seen as crucial for ongoing learning, hearing diverse perspectives, and maintaining the strong relationships and trust built during the fellowship. Poonam Parihar expressed a personal commitment to continuing these conversations, and other participants strongly agreed on the value of regular group discussions for staying updated.