Bhaskar Sawant is an AI architect and cybersecurity innovator with over 15 years of experience building intelligent, adaptive defense systems. He specializes in blending deep software engineering with machine learning to create solutions that evolve in real-time and is an IEEE Senior Member advocating for responsible AI.

In this episode, Bhaskar delivers a powerful presentation on why "security isn't a tool, it's a habit," arguing that millions spent on technology can be undone by a single human error. He shares real-world examples of security failures caused not by technology, but by poor habits like sharing passwords or neglecting CI/CD pipeline updates. Bhaskar outlines core principles for building a security culture, such as least privilege, making secure behavior the easiest option, and fostering transparency. He recounts his personal journey from a .NET developer to an AI security expert, driven by the increasing sophistication of cyberattacks.

Bhaskar also presents two case studies from his work. First, he describes how his team used machine learning to transform PowerShell threat detection from a noisy, reactive system into a proactive one, reducing false positives by 80% and automatically stopping ransomware. Second, he explains how they implemented observability-driven security in a large .NET Core application, reducing mean-time-to-detect from hours to minutes by unifying performance and security data. Finally, Bhaskar discusses the future of AI in cybersecurity, predicting a shift towards embedded and explainable AI, and defines innovation as improving existing systems to make them more secure.

In this episode, you’ll discover:

· Why security is a habit and culture, not just a set of tools.

· How poor human habits can undo millions of dollars in security investments.

· Core principles for building a security culture: least privilege and making security easy.

· A case study on using machine learning to reduce PowerShell alert noise by 80%.

· How AI-based threat detection automatically stopped ransomware before it spread.

· A case study on implementing observability-driven security to reduce detection time from hours to minutes.

· The importance of unifying performance and security data for real-time defense.

· The future of AI in cybersecurity: embedded, explainable, and guided copilots.

· Bhaskar's definition of innovation as improving and securing existing systems.

Connect With Bhaskar Sawant:

· LinkedIn: https://www.linkedin.com/in/bhaskar-bharat-sawant-533218122/

Chapters:

00:00 Welcome Bhaskar Sawant: AI Architect & Cybersecurity Innovator

01:13 Presentation: Security Isn't a Tool, It's a Habit

04:29 Real-World Examples: How Poor Habits Break Strong Security

07:04 Core Principles: Least Privilege and Making Security Easy

09:37 The Importance of Detection, Response, and Transparent Culture

10:43 Bhaskar's Journey: From .NET Developer to AI Security Expert

12:37 Habits an Organization Must Foster for Security

14:27 AI in the Game of Attack and Defense: Who Benefits More?

17:05 Case Study 1: Transforming PowerShell Threat Detection with Machine Learning

24:52 Case Study 2: Implementing Observability-Driven Security in a .NET Core System

29:57 The Future of AI in Cybersecurity: Embedded and Explainable AI

31:04 Hope for a Secure Digital World with AI

32:54 Innovation Defined: Improving and Securing Existing Systems

33:37 Connect with Bhaskar Sawant on LinkedIn

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

James Lang is the Managing Partner at OverLang Venture Partners and a former MedTech COO who scaled his previous startup to over $20 million in revenue and a team of 60 global employees. Now, he pioneers a sustainable, human-centered approach to AI, helping businesses integrate intelligence without destroying culture.

In this episode, James recounts his journey from a MedTech exit and a personal health crisis to reconnecting with a childhood friend to co-found OverLang. He details how their frustration with expensive, off-the-shelf AI solutions led them to build their own proprietary infrastructure, dramatically reducing compute costs from $800 to $8 per test. James shares his core philosophy on building teams, emphasizing servant leadership and creating a culture where high performers are eager to follow him to new ventures. He debunks the hype around "unsustainable AI" consultants who sell overpriced, generic solutions that often alienate employees and compromise data security.

James explains why he believes 90% of AI startups are doomed to fail, citing a lack of proper data vectoring, model selection, and an unsustainable focus on top-line revenue over profit. He offers practical advice for technical founders on building operational scaffolding, such as hiring for weaknesses and providing equity to align long-term incentives. James also discusses the critical importance of building feedback loops into AI systems and argues that AI should be a value multiplier for human teams, not a replacement. Finally, he defines innovation as removing friction to create useful solutions and urges businesses to be cautious about AI hype that ignores culture and sustainability.

In this episode, you’ll discover:

· James' journey from scaling a $20M MedTech company to founding an AI venture firm.

· The story behind OverLang's proprietary AI infrastructure and its massive cost savings.

· Core principles for building and scaling high-performance teams through servant leadership.

· How to identify and avoid unsustainable AI consultants and their overpriced solutions.

· Why 90% of AI startups fail due to poor data practices and a "revenue-first" mindset.

· Practical advice for technical founders on hiring for weaknesses and using equity.

· The critical role of human feedback loops in training and improving AI systems.

· Why AI should be a "value multiplier" for teams, not a replacement for people.

· James' definition of innovation as removing friction to create useful solutions.

Connect With James Lang & OverLang Venture Partners:

· Website & AI Playground: https://www.overlang.com/

· LinkedIn: https://www.linkedin.com/in/james-lang-94329271/

Chapters:

00:00 Welcome James Lang: MedTech COO to AI Venture Partner

01:28 From Health Crisis to AI Innovation: The Origin of OverLang

04:58 Reducing AI Compute Costs from $800 to $8: Building Proprietary Infrastructure

07:47 Scaling a Team of 60: Servant Leadership and Culture

11:16 Sustainable AI vs. "Butthole" Consultants: Avoiding the Hype

16:15 Why 90% of AI Startups Fail: Data Vectoring and Unsustainable Economics

18:33 Efficient Recruiting in the Zoom Era: Utilizing Platforms like Upwork

21:04 Operational Advice for Technical Founders: Hiring for Weaknesses & Equity

24:30 The Future of AI: The Importance of Automated Feedback Loops

26:36 AI as a Team Multiplier, Not a Replacement

30:03 Innovation Defined: Removing Friction to Create Useful Solutions

31:50 Connect with James Lang & Explore the OverLang AI Playground

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

Chris Majer is the founder and CEO of the Human Potential Project and the architect of the Redline Consulting Framework. With a background spanning from training US Marines and Special Forces to leading billion-dollar cultural transformations at companies like Microsoft and Intel, he now focuses on integrating AI as a strategic partner in leadership and business strategy. His mission has always been to explore the boundaries of human potential.

In this episode, Chris explains the necessary mindset shift for leaders to partner with AI, drawing historical parallels to the early days of radio, TV, and the internet, suggesting we are currently only using AI for what we already know how to do, not its full potential. He introduces his Redline Consulting Framework, which uses AI trained on his 35-year methodology to rapidly diagnose organizational blind spots, dramatically reducing assessment time from weeks to minutes. Chris also identifies a flawed understanding of learning as the single most important principle of human performance that corporate teams overlook. He distinguishes between "understanding" (a mental process) and "embodied competence" (the ability to take action without thinking), arguing that only practice and discomfort can bridge that gap.

Chris discusses the cultural barriers preventing large organizations from innovating with AI, tracing it back to industrial-era management practices built on distrust. He shares a tangible practice for building resilience called "centering" and emphasizes the need for leaders to transition from managers to guardians of the enterprise's mood and architects of its future. Finally, he argues that while AI excels at rote tasks, the uniquely human capacity to connect, feel, and truly lead is irreplaceable, and defines innovation not as invention, but as the incremental or radical improvement of existing ideas within a culture that tolerates risk.

In this episode, you’ll discover:

· The historical context for the current stage of AI adoption.

· How Chris's Redline Framework uses AI to rapidly diagnose organizational issues.

· The difference between understanding and "embodied competence" in human performance.

· Why practice and discomfort are essential for true learning and transformation.

· The industrial-era management practices holding back AI innovation in corporations.

· A somatic "centering" practice for building resilience and presence.

· The two fundamental responsibilities of a leader: guardian of mood and architect of future.

· Why AI will never replace the uniquely human capacity for connection and leadership.

· Chris's distinction between invention and innovation.

Connect With Chris Majer & Human Potential Project:

· Website: http://humanpotentialproject.com/

· Speaker Website: https://chrismajer.com/

· Book: "The Power to Transform" on Amazon

Chapters:

00:00 Welcome Chris Majer: Founder of Human Potential Project

01:28 A Divine Discontent: Exploring the Boundaries of Human Potential

02:30 Partnering with AI: A Historical Perspective on New Technology

06:34 The Redline Consulting Framework: Using AI for Organizational Diagnosis

08:27 The Missing Performance Principle: Embodied Competence vs. Understanding

11:20 The Power of Practice and the Discomfort of Learning

13:12 Industrial-Era Management: The Barrier to AI Innovation

16:34 Building Resilience: The Practice of "Centering"

17:48 The Two True Roles of a Leader

19:14 The Future of Leadership: Listening and Anticipation

20:04 AI and the Irreplaceable Human Potential for Connection

21:12 Career Advice for Students: Commitment and Lifelong Learning

22:46 Innovation Defined: Improvement Within a Risk-Tolerant Culture

24:08 Connect with Chris Majer and the Human Potential Project

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Faiz Gouri is a Lead AI Engineer at Microsoft, specializing in AI-driven cybersecurity for large-scale cloud infrastructure. He works on real-time anomaly detection and threat mitigation and is an active IEEE researcher with influential papers on optimizing distributed systems. His work sits at the critical intersection of cutting-edge AI and enterprise-level security.

In this episode, Faiz explains how his role at Microsoft organically combined AI, machine learning, and cybersecurity, emphasizing that security is a built-in feature, not an enhancement. He breaks down the high-level architecture of a real-time AI anomaly detection system, using a practical example of detecting suspicious login attempts from different locations. Faiz discusses his IEEE research on adaptive indexing, where machine learning dynamically adjusts database indexing based on query patterns, leading to significant performance improvements in distributed systems handling petabytes of data. He also explores the balance between deterministic rules and probabilistic AI in high-stakes cybersecurity, noting the critical importance of real-time threat mitigation to prevent costly downtime.

Faiz identifies data privacy in training models as a major pitfall for data scientists, using the example of AI chatbots in healthcare. For students and early-career professionals, he recommends cultivating intense curiosity and a commitment to continuous learning above any single technical skill. Reflecting on his own career, Faiz shares the advice to "take notes of everything" to retain knowledge in a rapidly evolving field. He expresses hope for AI's potential to create a more secure digital future, provided that security is core to its design. Finally, Faiz defines innovation as building something novel to solve a problem, often by drawing inspiration from and combining existing systems in unique ways.

In this episode, you’ll discover:

· How AI, machine learning, and cybersecurity converge in large-scale cloud infrastructure.

· The high-level architecture of real-time AI anomaly detection systems.

· How adaptive indexing with machine learning optimizes distributed databases.

· The balance between rule-based systems and AI in high-stakes security.

· The critical challenge of data privacy when training AI models with sensitive information.

· Why curiosity and continuous learning are the most important skills for an AI career.

· The career-defining advice to "take notes of everything" to retain knowledge.

· The potential of AI to create a more secure digital future.

· Faiz's definition of innovation as building novel solutions inspired by existing systems.

Connect With Faiz Gouri:

· LinkedIn: https://www.linkedin.com/in/faizgouri/

Chapters:

00:00 Welcome Faiz Gouri: Lead AI Engineer at Microsoft

01:51 The Intersection of AI, Machine Learning, and Cybersecurity at Microsoft

04:13 Architecture of a Real-Time AI Anomaly Detection System

07:08 IEEE Research: Adaptive Indexing for Distributed Systems

09:43 Balancing AI and Rule-Based Systems in High-Stakes Security

11:34 AI vs. Traditional Systems in Threat Detection

12:38 Major Pitfall: Data Privacy in Training Production AI Models

15:28 Foundational Skill for an AI Career: Curiosity and Continuous Learning

17:35 Career Advice: The Importance of Taking Notes

19:35 The Future of AI in Creating a Secure Digital World

20:37 Innovation Defined: Building Novel Solutions from Existing Inspiration

21:41 Connect with Faiz Gouri on LinkedIn

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

Mark Burnett is a former database programmer who transitioned from patient to inventor after being diagnosed with both Parkinson's and Alzheimer's diseases. Applying his problem-solving skills, he developed MyBrainRestore, a unique supplement using ultra-chill milled technology.

In this episode, Mark recounts his journey from a shocking dual diagnosis to becoming an inventor, driven by his refusal to accept a grim prognosis. He explains how Japanese research on a natural ingredient, which failed when synthesized for drug development, became his "aha" moment. Mark details his discovery that heat was destroying the active properties of the ingredient, leading him to pioneer his ultra-chill milling process. He shares his personal success story, including running a marathon and documenting a 10% reversal in his Alzheimer's blood test biomarkers—a result his doctor called impossible.

Mark discusses the human-centered design of his product as a powder, making it easier for patients and caregivers to use. He explains how MyBrainRestore targets the brain's glymphatic system to clear out build-up and encourages patients to track their own symptoms to measure progress. He also reveals that some of his customers are doctors taking the product themselves. Finally, Mark shares the most important lesson he learned from transitioning from code to biology: the need to think outside the box when medical science offers no immediate solutions.

In this episode, you’ll discover:

· Mark's journey from database programmer to Parkinson's & Alzheimer's patient and inventor.

· How Japanese research on a natural ingredient inspired his own discovery.

· The problem with heat in supplement processing and the "Ultra-Chill Milled" solution.

· Mark's personal results: running a marathon and reversing Alzheimer's biomarkers.

· Why MyBrainRestore is designed as a powder for easier consumption by patients and caregivers.

· The role of the brain's glymphatic system in neurological health.

· How patients can track their own progress by journaling symptoms like tripping or memory loss.

· Mark's experience with doctors who are skeptical but also buying the product for themselves.

· The importance of thinking outside the box when faced with an incurable diagnosis.

· Mark's definition of innovation as looking at problems in new ways.

Connect With Mark Burnett & MyBrainRestore:

· Website: https://mybrainrestore.com/

· Phone: 202-536-5000 (Answers 7 days a week)

Chapters:

00:00 Welcome Mark Burnett: From Database Programmer to Inventor

01:17 The Parkinson's & Alzheimer's Diagnosis and the Decision to Engineer a Solution

03:34 Discovering the Japanese Research: The "Aha Moment" for a Natural Cure

08:24 The Heat Problem: How "Ultra-Chill Milled" Technology was Born

10:55 Documenting Reversal: Running a Marathon & Improving Alzheimer's Biomarkers by 10%

16:23 User-Centered Design: Why MyBrainRestore is a Powder, Not a Pill

21:09 Tracking Progress at Home: Journaling Symptoms for Evidence

23:44 Doctors as Customers: Skepticism and Personal Use

25:04 Transitioning from Code to Biology: The Need to Think Outside the Box

30:50 Innovation Defined: Looking at Problems in New Ways

35:06 Connect with Mark Burnett & Get the MyBrainRestore E-book

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

Mia Umanos is the Co-CEO of Clickvoyant, a Techstars-backed AI company that turns messy marketing data into clear, actionable insights for e-commerce and retail brands. She started her career in journalism and PR before leveraging her narrative skills to build AI tools that focus on the human decisions behind the numbers.

In this episode, Mia shares her pivotal career shift from faxing press releases in public relations to becoming obsessed with digital analytics. She explains how Clickvoyant addresses the challenges of traditional data science by building an AI agent that provides specific, "do-this-now" advice rather than confusing dashboards. Mia discusses the core technical challenge of integrating disparate data sources and her bold prediction that conversational AI will replace dashboards as the primary way people interact with analytics. She shares a real-world example where AI surfaced a counter-intuitive but highly valuable customer segment that human analysts had overlooked, leading to a double in conversion rates.

Mia also addresses the biggest mistake tech teams make when building user-friendly analytics, emphasizing the need for simplicity over complexity. She argues that human empathy remains irreplaceable in AI-driven marketing, as it's essential for building trust and making sales. Mia discusses her podcast, "AI Kid You Not," co-hosted with her 14-year-old daughter, and the humbling lessons she's learned about AI from their conversations. Finally, she explains how parenting has taught her the art of "deep listening" to uncover the true questions behind client data requests and defines innovation as creatively responding to a moment to solve problems in new ways.

In this episode, you’ll discover:

· Mia's journey from journalism and PR to AI-powered marketing analytics.

· How Clickvoyant's AI turns messy marketing data into clear, actionable advice.

· Why dashboards are becoming obsolete and will be replaced by conversational AI.

· A real-world case study: How AI discovered a hidden, high-value customer segment.

· The biggest mistake tech teams make when building user-facing analytics.

· Why human empathy is irreplaceable, even in an AI-driven world.

· Lessons on AI and the future of work from Mia's podcast with her teenage daughter.

· How parenting skills like "deep listening" apply to data consulting.

· Mia's advice for business leaders: move beyond simple KPIs and embrace data complexity.

· Mia's definition of innovation: creatively responding to a moment to solve problems.

Connect With Mia Umanos & Clickvoyant:

· Clickvoyant Website: https://clickvoyant.com/

· Podcast: AI Kid You Not (available on major platforms)

Chapters:

00:00 Welcome Mia Umanos: Co-CEO of Clickvoyant

01:15 From Journalism & PR to a Passion for Data Analytics

03:58 Building Clickvoyant: An AI Agent for Actionable Marketing Insights

05:47 The Technical Challenge: Integrating Messy, Disparate Marketing Data

08:16 The End of Dashboards: Why Conversational AI is the Future of Analytics

11:04 Human-AI Collaboration: How AI Surfaced a Counter-Intuitive Customer Segment

13:30 The Biggest Mistake Tech Teams Make: Building Too Much Complexity

15:37 The Irreplaceable Human Skill in an AI World: Empathy

18:07 Lessons from "AI Kid You Not": A Podcast with Her Teenage Daughter

20:41 Applying Parenting Skills to Data Consulting: The Power of Deep Listening

24:14 Advice for Business Leaders: Move Beyond KPIs to Complex Data Thinking

25:37 Innovation Defined: Creatively Responding to Solve Problems

27:39 Connect with Mia Umanos & Try Clickvoyant

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

Michele Novack is the founder of Cardinals Byte and a cybersecurity expert with a unique mission to make digital protection accessible for small businesses and professionals. After a 30-year career in the financial sector, she pivoted to help accountants, lawyers, and other regulated industries navigate compliance and risk without the overwhelming technical jargon. She is also an author who focuses on the powerful connection between mindset and security.

In this episode, Michele shares her pivotal journey from a long career in finance to becoming a cybersecurity engineer in her fifties, driven by the realization that her role was becoming obsolete. She provides an overview of Cardinals Byte, her firm that helps small businesses—particularly in regulated fields like finance and law—manage cybersecurity risk and compliance through assessments, monitoring, and training. Michele addresses the biggest misconception small business owners have: that having an "IT guy" and antivirus software is enough to protect them. She explains her process of translating complex compliance regulations into simple, actionable steps, using risk assessments with clear explanations and free training resources.

Michele recommends vulnerability scanning as a non-negotiable, low-cost cybersecurity practice for small businesses and highlights free government resources for protection. She discusses how her philosophy of "living fully" influences her approach to risk management, emphasizing resilience, continuous learning, and the importance of building trust with clients. Michele shares a story of overcome a client's significant skepticism by using a powerful visual metaphor to establish trust. She identifies AI and deepfakes as the most concerning emerging cyber threats and urges small businesses to educate themselves. Finally, she offers advice on bridging the communication gap between technical professionals and business owners and defines innovation as the process of holistically understanding a problem and creating a future-forward solution.

In this episode, you’ll discover:

· Michele's pivot from a 30-year finance career to cybersecurity in her fifties.

· How Cardinals Byte helps small businesses navigate cybersecurity compliance without jargon.

· The common misconception that having an "IT guy" is enough for cybersecurity.

· How Michele translates complex regulations into simple, actionable steps.

· Vulnerability scanning as a critical, low-cost security practice for small businesses.

· Free government resources available for business cybersecurity.

· How a "live fully" mindset influences risk management and client trust.

· The emerging threat of AI and deepfakes to small businesses.

· Bridging the communication gap between technical experts and business owners.

· Michele's definition of innovation as future-forward problem-solving.

Connect With Michele Novack & Cardinals Byte:

· Website: https://www.cardinalsbytes.com/

· LinkedIn: https://www.linkedin.com/company/cardinalsbyte/?viewAsMember=true

Chapters:

00:00 Welcome Michele Novack: Founder of Cardinals Byte

01:15 From Finance Leadership to Cybersecurity Engineer at 50+

03:12 Cardinals Byte Overview: Helping Small Business with Risk & Compliance

05:14 The Biggest Misconception: "I Have an IT Guy" Isn't Enough

08:20 Translating Complex Compliance into Simple Steps for Business Owners

10:20 One Non-Negotiable Cyber Practice: Vulnerability Scanning

12:38 Applying a "Live Fully" Mindset to Risk Management

17:04 Overcoming Client Skepticism and Building Trust: The Visual Metaphor

21:01 What a "Win" Looks Like: Client Protection and Peace of Mind

22:25 Emerging Cyber Threats: The Rise of AI and Deepfakes

25:18 Bridging the Communication Gap Between Tech Experts and Business Owners

29:06 Innovation Defined: Holistic, Future-Forward Problem Solving

32:35 Connect with Michele Novack and Get a Free Risk Assessment

Lin Wu arrived in the United States alone at age 16, fleeing a revolution with only $600 in his pocket. From working in a broom shop to survive, he went on to become a pioneering engineer in the golden era of Silicon Valley, helping architect supercomputers that challenged IBM and eventually founding a semiconductor company acquired by Toshiba.

In this episode, Lin shares his incredible firsthand account of the valley's formative years—a time of physical hardware, grueling 12-hour days, and audacious innovation. He discusses his journey from IBM to becoming one of the first 18 engineers at Amdahl Corporation, where they built the world's fastest mainframe computer. Lin also details the process of founding his own custom chip design company, building a relationship with Toshiba, and navigating a successful acquisition.

This is a masterclass in resilience, resourcefulness, and raw technical ingenuity. Lin offers timeless advice for young engineers and entrepreneurs on the importance of market validation over just building "better mousetraps," and reflects on the similarities between the early supercomputer race and today's AI boom.

Key Topics Discussed:

* Survival Mode: How working as a broom stitcher as a teenager taught Lin self-sufficiency.

* The Amdahl Story: Inside the intense race to build a supercomputer 2.5x faster than IBM in the 1970s.

* Founding & Exit: The journey of starting Integrated CMOS Systems and the process of being acquired by Toshiba.

* Entrepreneurial Advice: Why knowing your market is more critical than just loving your product technology.

* Defining Innovation: Moving beyond invention to practical, marketable usefulness.

Resources Mentioned:

Lin Wu’s Memoir: "Destination Silicon Valley: A Memoir - From Broom Stitcher to Supercomputer Engineer" - Available on Amazon: https://www.amazon.com/Destination-Silicon-Valley-Broom-Supercomputer-ebook/dp/B0FVGF1BKR

Chapters:

00:00 Intro & Cold Open

01:18 Surviving at 16: The Broom Shop Story

04:48 From Ivy League to IBM

06:59 The Dawn of Supercomputers: Joining Amdahl Corporation

08:43 The Immigrant Perspective in Early Tech

10:18 Founding Integrated CMOS Systems

12:53 The Toshiba Acquisition Process

15:49 Frugal Advice for Young Professionals: Know Your Market

17:34 Grit, Ambition, and Disrupting the Old

18:48 First Moves for Outsider Entrepreneurs

20:19 Early Supercomputers vs. Today's AI Boom

21:57 Defining True Innovation

23:09 "Destination Silicon Valley": Lin Wu's Memoir

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

Sachin Hiriyanna is a Staff Software Engineer at Navan, where he leads major platform initiatives for their Android app used by over 100,000 people. His career spans from co-founding his own startup to refactoring core systems at scale, operating at the intersection of deep technical work and strategic leadership.

In this episode, Sachin recounts his journey from a curious childhood tinkerer to a staff engineer, including his early experiences with a startup and developing a life-saving app for opioid overdose victims. He dives into the realities of large-scale refactoring, sharing the detailed process of migrating Navan's Android app from the legacy Object Box database to modern solutions like Room. Sachin explains his methodical, four-phase approach to this complex migration, emphasizing the importance of planning, communication, and using end-to-end tests to ensure stability without disrupting feature development. He discusses the role of a Staff Engineer, highlighting that it's less about just writing brilliant code and more about proactive problem discovery, cross-team collaboration, and shaping technical strategy—illustrated by his team's proactive migration to support Android's new 16KB page size requirement.

Sachin also reflects on his time as a startup founder and how that entrepreneurial mindset influences his work in corporate platform engineering. He defines innovation as the process of discovering a problem and creating an elegant, scalable, and applicable solution. Finally, he offers advice for engineers aspiring to reach staff level, stressing the need to look beyond well-defined tasks and take initiative in identifying and solving broader product challenges.

In this episode, you’ll discover:

· Sachin's journey from startup founder to Staff Software Engineer at Navan.

· The process of large-scale refactoring and migrating legacy databases (Object Box to Room).

· A four-phase approach to complex technical migrations without disrupting product delivery.

· The role of a Staff Engineer: proactive problem discovery and cross-team strategy.

· A real-world example of proactive engineering: migrating to Android's 16KB page size.

· How an entrepreneurial background benefits corporate platform engineering.

· Sachin's definition of innovation: elegant, scalable solutions to discovered problems.

· Advice for senior engineers aspiring to reach the staff level.

· The importance of end-to-end testing during major code refactoring.

Connect With Sachin Hiriyanna:

· LinkedIn: https://www.linkedin.com/in/hsachin/

Chapters:

00:00 Welcome Sachin Hiriyanna: Staff Software Engineer at Navan

01:10 From Childhood Tinkerer to Startup Founder & Staff Engineer

05:01 Tech for Good: Developing an App for Opioid Overdose Victims

08:19 Startup Experience: Building a Fintech App for Price Protection

11:26 Leading Large-Scale Refactoring: Migrating Databases at Navan

16:47 A Four-Phase Approach to Complex Technical Migrations

23:02 The Staff Engineer Role: Proactive Strategy vs. Just Brilliant Code

26:46 Proactive Problem Solving: The Android 16KB Page Size Migration

28:11 Innovation Defined: Elegant, Scalable Solutions to Discovered Problems

31:10 Connect with Sachin Hiriyanna on LinkedIn

Support the Show:

· Fuel the podcast: https://iferia.nestuge.com/supportme

· Subscribe and leave a review!

· Share

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast

Vlad Savinov is a 24-year-old AI Team Lead specializing in large language model (LLM) pre-training. He manages the complex infrastructure required to train models with billions of parameters, a journey that began with a university fascination with AI and reinforcement learning.

In this episode, Vlad demystifies the technical reality of large language models, explaining them not as "black boxes" but as vast collections of numbers and mathematical operations. He uses a brain analogy to describe parameters but emphasizes their true nature as numerical weights. Vlad pulls back the curtain on the distributed training infrastructure needed to handle these massive models, explaining how the computation is split across an "orchestra" of GPUs because no single device has enough memory. He describes his team's "war room" during a training run, where the focus is on making experiments and the final run as fast and smooth as possible to deliver high-quality models quickly.

Vlad discusses the strategic reasons why companies open-source expensive models like DeepSeek, citing benefits like attracting talent, community contributions, and building brand reputation. He debunks the misconception that AI is an "out-of-the-box" automation solution, highlighting the significant work required in data cleaning and evaluation. Vlad shares the challenges of large-scale training, such as training divergence, and expresses excitement about recent breakthroughs like lower precision training that speed up the process. He offers advice for students aspiring to work in AI, emphasizing the importance of building projects and implementing ideas from papers. Finally, Vlad addresses the AI safety debate, advocating for a balanced approach that considers both acceleration and necessary precautions, and defines innovation as the hard work of making a theoretical idea work in the real world for the first time.

In this episode, you’ll discover:

· What a parameter actually is in a large language model (it's just a number).

· How distributed training infrastructure splits massive models across multiple GPUs.

· The strategic reasons why companies open-source expensive AI models.

· The biggest misconception about AI: it's not an "out-of-the-box" automation solution.

· Challenges in large-scale training, like training divergence and debugging.

· Exciting technical breakthroughs, such as lower precision training for efficiency.

· Advice for students: build projects and implement ideas from research papers.

· A balanced perspective on the AI safety debate (acceleration vs. caution).

· Vlad's definition of innovation: making a theoretical idea work in the real world.

Connect With Vlad Savinov:

· LinkedIn: https://www.linkedin.com/in/vladislav-savinov/

Chapters:

00:00 Welcome Vlad Savinov: AI Team Lead & LLM Pre-training Expert

01:07 From Math Competitions to AI: The Journey Began in University

03:01 Demystifying LLMs: Parameters are Just Numbers, Not Black Boxes

04:56 Distributed Training Infrastructure: An Orchestra of GPUs

07:01 The "War Room": Optimizing Training Speed and Quality

08:31 Teaching Algorithms and Its Impact on Engineering Leadership

10:38 The Strategic Value of Open-Sourcing Expensive AI Models

13:01 The Biggest Misconception About AI: It's Not "Out-of-the-Box"

15:09 Challenges in Large-Scale Training: Divergence and Debugging

17:31 Exciting Breakthroughs: Lower Precision Training for Efficiency

19:36 Advice for Students: Build Projects and Implement Research Papers

21:39 The AI Safety Debate: Balancing Acceleration and Caution

24:56 Leading Experienced Engineering Teams at a Young Age

26:54 Innovation Defined: Making Theoretical Ideas Work in the Real World

28:22 Connect with Vlad Savinov on LinkedIn

Want to Be a Guest on The Iferia TechCast?

· Reach out to Ezekiel on PodMatch

· PodMatch Host Profile: https://podmatch.com/hostdetailpreview/theiferiatechcast