OPEN Tech Talks: AI worth Talking| Artificial Intelligence |Tools & Tips

What is the business value of private AI, and why should companies consider owning the AI systems that connect to their internal data and operations? In Episode 202 of Open Tech Talks, host Kashif Manzoor speaks with James Lang, an operator, investor, and technology strategist at Overlang Venture Partners, about private AI, business growth, AI security, legal technology, business valuation, and the future of work.

James has a background in MedTech, operations, technology, and business growth. He now works with small and medium-sized businesses to help them improve operations and use technology in practical ways.

A key idea in this episode is AI ownership.

Many companies use public AI tools. They may use different tools for different tasks. However, their core business systems may still depend on external platforms.

James explains why companies may benefit from owning key AI systems that connect to their internal operations and data. These systems can run in a private environment and support business processes.

The goal is not to build every AI system from scratch.

The goal is to own the AI systems that are important to the business.

Episode # 202

James Lang is a Venture Partner and former MedTech COO who has helped generate more than $20 million in revenue while building and leading a 60‑person global operations team.

• Website: OverLang

What Listeners Will Learn:

• Why Companies Should Own Their Core AI Systems
• How Private AI Can Protect Business Data
• How AI Can Increase Business Value
• How AI Can Support Business Growth
• Why AI Ownership Can Matter During an Exit
• How Companies Can Combine Private and External AI
• How AI Is Changing Legal Services
• Why Human Review Still Matters
• How AI Can Improve Legal Research and Document Work
• How AI Is Changing the Value of Human Work
• Why Information Filtering Is Becoming More Important
• How Businesses Can Approach AI Adoption
• Why AI Governance Matters
• How Advanced AI Could Create New Business Risks
• How Companies Can Balance AI Innovation and Security

• OverLang

Insights on agent vulnerabilities and safety measures in AI.

In this week's briefing, we explore critical insights on AI agents and their security vulnerabilities. Kashif Manzoor discusses recent incidents that highlight the importance of understanding and controlling AI agents' capabilities.

Understanding AI Agent Vulnerabilities Kashif opens by reflecting on his journey documenting AI developments over the past two years. The discussion transitions to alarming news about AI agents escaping their designated sandboxes, particularly an OpenAI research agent that inadvertently accessed the internet. This incident occurred on September 20th, when the agent used its DNS resolver to send queries to a public chatbot, raising concerns about the security measures in place.

This incident reminds us that our AI agents may have uncharted paths that could be exploited.

Kashif emphasizes the necessity for organizations to evaluate how quickly they can respond to similar incidents and to map the potential reach of their AI agents.

The Intersection of AI Safety and Diplomacy The conversation also explores recent engagements between AI CEOs and the UN Security Council, where discussions centered on the need for global standards for testing AI models. The potential geopolitical implications of AI safety are paramount as nations like the US and China begin to navigate these complexities.

Looking Ahead to AI's Future As AI continues to evolve, Kashif highlights innovations like the Stanford researchers' Home Bud Body, showcasing how AI can autonomously navigate and learn from new environments. This shift in capability challenges leaders to rethink which tasks AI can handle in various settings.

Final Thoughts In closing, Kashif prompts readers to reflect on their AI agents' reach and capabilities. What paths might they have that you don't know about? This introspection is vital for ensuring the safety and integrity of AI systems in an increasingly automated world.

How can businesses use AI agents to automate operations, improve productivity, preserve organizational knowledge, and help employees become more valuable? In Episode 200 of Open Tech Talks, host Kashif Manzoor speaks with Nathan Graham, AI consultant, educator, and AI systems builder, about the rise of agentic AI, AI-powered C-suites, business automation, human-AI collaboration, and the changing future of work.

Nathan works with businesses to design and implement AI systems that can support functions across the organization, including marketing, finance, operations and administration. Rather than viewing AI as a tool to replace employees, he focuses on using it to work alongside people and automate the repetitive work that consumes their time.

One key concept in this episode is the AI-powered C-suite. Nathan explains how businesses can build AI systems that work alongside human executives and employees, helping manage workflows while staying aligned with company-specific governance and priorities.

The conversation also introduces the idea of a company brain. When an employee leaves an organization, valuable operational knowledge can often disappear. AI can help preserve workflows, processes, and organizational knowledge so new employees don't necessarily have to start from zero.

Nathan also explains his concept of a learning fabric, where proven workflows can potentially be shared across participating AI systems while keeping company-specific information and competitive strategies separate. The goal is to allow businesses to benefit from improved processes without exposing sensitive business data.

Episode # 200

Nathan Graham, AI Consultant, Educator, and AI Systems Builder, Synthetic Echo

He is the author of The Synthetic Echo and AI Workflows: From Cold Lead to Closed Deal. He is also the author of Why Won't He Just Log Off?, which explores gaming, parenting, human needs, and technology's growing impact on family relationships.

• Website: Synthetic Echo

What Listeners Will Learn:

• How AI Agents Can Transform Business Operations
• What an AI-Powered C-Suite Looks Like
• How a Company Brain Can Preserve Organizational Knowledge
• How Learning Fabrics Can Improve AI Workflows
• Where AI Automation Creates the Most Value
• Where Businesses Should Keep Humans in the Loop
• Why Human Connection Still Matters
• How AI Can Increase Employee Productivity
• How AI Could Change the Future of Jobs
• How to Become More Valuable With AI
• How Businesses Should Approach AI Adoption
• Why Too Many AI Tools Can Become a Problem
• How Agentic AI Could Replace Traditional Dashboards
• How AI Agents Could Change SaaS Applications
• How to Start Learning AI Today

• Synthetic Echo

How can developers use multiple AI models without constantly switching tools or hitting rate limits? In Episode 199 of Open Tech Talks, Kashif Manzoor speaks with Jonathan Archer, co-founder of OpenLLM, about AI gateways, Bring Your Own Key (BYOK), multi-model AI workflows, AI coding agents, and the future of developer productivity.

As developers increasingly use Claude, ChatGPT, Grok, Gemini, and other AI models, managing multiple subscriptions, API keys, usage limits, and development environments is becoming a real challenge. Jonathan explains how an AI gateway can provide a unified endpoint for multiple AI providers, letting developers route requests between models and create fallback workflows when a provider hits a usage limit or goes down.

The conversation also explores why multi-model AI development could become increasingly important. Instead of relying on a single AI provider throughout an entire workflow, developers can assign different models to different tasks such as planning, coding, reviewing, and fixing. Jonathan shares how this approach can improve results while reducing costs and avoiding blind spots that can occur when the same model evaluates its own work.

Kashif and Jonathan also discuss AI-assisted software development and graph engineering, where multiple AI agents work together in structured loops to solve problems. They explore how these workflows are changing the way developers build software and how AI coding tools can enable very small teams to develop sophisticated products.

Beyond technology, Jonathan shares practical lessons from building and launching OpenLLM. The discussion covers customer discovery, product development, workshops, cold outreach, social media marketing, AI-generated content, and the challenges solo founders and small startup teams face.

The episode also covers security and privacy, including how AI gateways handle credentials, why local processing can matter, and why developers should pay close attention to AI providers' terms of service when using subscription-based AI access.

The episode concludes with advice for beginners who want to learn coding and build their own products using AI coding tools. Jonathan's message is straightforward: start experimenting, build something small, learn by doing, and become someone who uses AI rather than someone who fears it.

Episode # 199

His work focuses on improving the developer experience around AI models, reducing rate-limit interruptions, and enabling developers to combine different models for coding and agentic workflows.

• Website: OpenLLM

What Listeners Will Learn:

• How Bring Your Own Key AI Works
• Why Developers Use Multiple AI Subscriptions
• How AI Gateways Reduce Workflow Interruptions
• How Fallback Chains Work Across AI Models
• Why Multi-Model Agent Workflows Can Produce Better Results
• How Different AI Models Can Complement Each Other
• How AI Coding Agents Are Changing Software Development
• How Developers Can Build AI Products With Small Teams
• How Customer Feedback Drives AI Product Development
• How to Market an Early-Stage AI Product
• Why Security and Terms of Service Matter in AI Gateways
• How Beginners Can Start Building With AI Coding Tools
• Why AI Skills Are Becoming Essential for Developers
• How to Start Learning AI-Assisted Coding Today

• OpenLLM

In Episode 196 of Open Tech Talks, host Kashif Manzoor speaks with Sebastian Wernicke, a data and AI expert with 20 years of experience working across data science, analytics, machine learning, and artificial intelligence.

The conversation explores how organizations should approach AI transformation without getting distracted by the latest technology trends.

Sebastian shares a powerful principle that has guided his work:

Purpose first, data second.

Rather than starting with a new AI model or technology and looking for somewhere to use it, organizations should first understand the business problem they are trying to solve.

The discussion draws on Sebastian's experience working with organizations on complex data and analytics projects, including a logistics optimization project where the technical solution could deliver significant fuel savings but encountered unexpected organizational and operational challenges.

The conversation highlights an important lesson for data scientists and AI teams: a technically successful model does not automatically translate into business success.

Kashif and Sebastian also discuss the challenges of data preparation, data silos, institutional knowledge, and the importance of understanding how data is actually generated and used inside an organization.

Sebastian explains why spending time with business teams and even visiting operational environments can reveal details that are invisible in datasets.

The conversation then moves toward generative AI and the growing tendency for organizations to replace traditional data science and machine learning with the latest AI technologies.

Sebastian argues that organizations should choose the simplest and most appropriate technology for the problem, rather than automatically selecting an LLM or generative AI solution.

The episode also explores data platforms, data mesh, data governance, and the organizational challenges that technology alone cannot solve.

Episode # 198

For over 20 years, Sebastian has guided organizations worldwide to harness the power of data and AI to achieve breakthrough transformation.

• Website: DataInspired

What Listeners Will Learn:

1. Why Purpose Should Come Before Technology
2. How to Identify the Right Business Problem
3. How Data Creates Business Value
4. Why AI Projects Fail Beyond the Algorithm
5. How to Overcome Data Silos
6. Why Data Preparation Still Matters
7. How to Connect AI Use Cases With Business Outcomes
8. Why Traditional Machine Learning Still Matters
9. When Generative AI Is the Right Choice
10. How to Build Better Data Foundations
11. Why Technology Cannot Fix Organizational Problems
12. How Technical Professionals Can Build Long-Term Careers

• DataInspired

In Episode 197 of Open Tech Talks, host Kashif Manzoor speaks with Francis Brero, VP of AI Strategy at HD Insights, about what it really means for an enterprise to become AI native.

Francis shares his experience helping an established enterprise organization move into the AI era and explains why AI transformation is much more than adding an LLM or purchasing the latest AI tools.

The conversation begins with a practical framework for identifying where AI can create value. Francis explains how organizations can break down jobs into smaller tasks, assess each task's value, and identify which activities are suitable for AI automation.

The discussion then moves into one of the biggest challenges facing enterprises today. Many organizations have an AI mandate, but they are adopting AI simply because they feel they need to "do AI." Francis explains why organizations need to focus on the problems they are trying to solve instead of chasing every new model and tool.

A major theme of the conversation is institutional knowledge. Enterprises often depend heavily on experienced employees who know how things work. Francis explains how AI can turn that knowledge into an organizational asset rather than letting it remain locked inside individual employees.

The episode also explores AI-native software development, multi-agent workflows, AI-powered code reviews, system optimization, security controls, and the role of different AI models in the software development lifecycle.

Francis also explains why traditional machine learning remains important. Not every problem needs an LLM, especially when organizations are working with structured data or systems that require deterministic outcomes.

The conversation concludes with a discussion about managing the incredible speed of AI innovation. Instead of changing systems every time a new model appears, organizations should build flexible systems that let them introduce new models and capabilities without rebuilding everything.

Episode # 197

Francis Brero is a longtime AI and GTM operator who has spent the past 15 years building data-driven SaaS products that help revenue teams make smarter decisions.

• Website: HDInsights
• LinkedIn: Francis Brero

What Listeners Will Learn:

• How to Build an AI Native Enterprise
• How to Identify the Right AI Use Cases
• How to Break Jobs Into AI-Ready Tasks
• How to Measure the Value of AI Automation
• How to Turn Institutional Knowledge Into an Organizational Asset
• How to Build AI Native Development Workflows
• How AI Agents Can Improve Code Reviews
• How to Use Multiple AI Models Effectively
• When Traditional Machine Learning Is Better Than LLMs
• How to Build AI Systems That Can Adapt to New Models
• How to Balance AI Innovation With Enterprise Stability
• Why System Thinking Matters in AI Transformation

• Website: HDInsights
• LinkedIn: Francis Brero

In Episode 196 of Open Tech Talks, host Kashif Manzoor sits down with Kate Marshall, a cybersecurity and AI adoption specialist, to discuss one of the biggest challenges facing organizations today:

How can companies move quickly with AI while keeping their data, people, and systems secure?

Kate brings nearly two decades of cybersecurity experience to the conversation and explains why AI adoption cannot be separated from cybersecurity, privacy, governance, and employee training.

The conversation starts with the growing adoption of AI inside organizations and the different approaches taken by highly regulated industries versus organizations that are aggressively adopting autonomous AI agents.

Kate explains that the biggest AI security risks often come from people using AI tools without understanding the risks.

From downloading untrusted skills and encountering prompt injection to connecting AI systems to internal data, seemingly harmless actions can create serious security problems.

The discussion then moves into practical steps organizations can take to adopt AI safely without creating unnecessary barriers.

Kate recommends starting with two basic foundations:

• An AI policy
• An approved AI tools list

But policy alone isn't enough.

Organizations must also train employees, define permissions, establish governance, create incident-response processes, and continuously improve their AI adoption strategy.

The episode also explores the importance of measuring AI ROI, creating digital twins for critical business processes, training employees to become effective AI operators, and ensuring that teams understand how to use AI tools responsibly.

For developers and product teams building applications with LLMs, Kate highlights another important issue: accountability.

When an AI-powered product makes a mistake, who is responsible?

That question needs to be considered during product design—not after something goes wrong.

The episode closes with advice for professionals who want to learn AI security, as well as a message for anyone feeling overwhelmed by the speed of AI development:

Episode # 196

She helps organizations adopt AI with strong governance, workforce readiness, and practical support. She spent 18 years at the SANS Institute, leading Summits creating rapid-response training for emerging cybersecurity threats, curriculum marketing strategy, and courseware development

• Website: KateMarshall

What Listeners Will Learn:

• How to Adopt AI Safely
• How to Build an AI Governance Framework
• How to Protect Sensitive Data When Using AI
• How to Manage AI Security Risks
• Why Employee AI Training Matters
• How to Balance AI Innovation and Security
• How to Create an Effective AI Policy
• How to Evaluate AI Tools and Permissions
• How to Build Secure AI Products
• How to Prepare Your Workforce for AI

• KateMarshall

Artificial Intelligence is no longer just for large enterprises with massive IT budgets. Today, even the smallest businesses can automate repetitive work, improve customer service, and reclaim valuable hours using AI.

Chapters:

00:00 Introduction to Anne Cantera 01:12 The Role of AI in Small Businesses 03:43 Identifying Problems in Small Businesses 06:41 Implementing AI Solutions 10:24 Tech Stack and Feasibility for Small Businesses 12:30 Driving Towards Workable AI Products 13:43 Adoption Challenges and Experiences 15:43 Measuring ROI in AI Implementations 18:07 Advice for Small Business Owners 19:40 Educating Yourself on AI 22:43 Public Perception of AI 24:09 Conclusion and Contact Information

Episode # 146

She is an AI and voice technology expert, conversation designer, and founder of Elementyl Intelligence

• Website: Elementyl Intelligence
• LinkedIn

What Listeners Will Learn:

• Why small businesses may adopt AI faster than large enterprises
• How to identify business problems that AI can solve
• A practical framework for assessing AI opportunities
• When AI is the right solution and when it isn't
• How voice agents can automate customer interactions
• The benefits of AI-powered inbox and workflow automation
• How custom AI solutions are becoming affordable for SMEs
• Why ROI should drive every AI investment
• The importance of AI literacy for business owners
• How to become a better AI operator.
• Why monitoring and continuous improvement matter after AI deployment.
• Common mistakes businesses make when implementing AI.
• How to avoid misinformation and AI hype.

• Elementyl Intelligence
• Voice of AI - Free AI Education Platform
• Elemental Intelligence - AI Consulting for Small Business
• Reuven Cohen - Trusted AI Educator

Artificial Intelligence has completely transformed software development. What started with autocomplete has evolved into AI coding agents capable of building entire applications from simple prompts. But as organizations adopt these tools, new challenges emerge: cost, security, vendor lock-in, enterprise governance, and the future role of software engineers.

Today's conversation isn't about one product. It's about understanding where software development is heading over the next few years.

Joining me is Emilie Schario, VP of Engineering at Kilo, where she works on agentic engineering and multi-model AI developer tools. We discuss how enterprises should think about AI coding assistants, why relying on a single AI model may not be the future, how engineering teams should prioritize product development, and what both junior and senior developers need to do to stay relevant.

Whether you're a developer, engineering leader, CTO, startup founder, or simply curious about AI transforming software engineering, this episode is packed with practical insights.

Let's get started.

Chapters:

00:00 Introduction to Emily Sherio and Kilo Code 04:13 The Role of Kilo in Developer Productivity 07:52 Choosing the Right AI Models for Development 12:24 Building Features that Matter 16:38 The Importance of Code Review and Feature Management 19:24 The Future of Development and AI Integration 25:59 The Necessity of Learning Coding Principles 30:16 Enhancing Security in Coding Tools

Episode # 194

She works on agentic engineering and multi-model AI developer tools

Guests Links:

• LinkedIn

What Listeners Will Learn:

• Why AI coding is moving toward multi-model architectures.
• The advantages and risks of proprietary AI coding tools.
• How organizations can reduce AI infrastructure costs.
• Why vendor lock-in is becoming an AI challenge.
• The future of developer productivity.
• Why code review is becoming more important than writing code.
• Product management lessons every startup should know.
• When to remove features rather than add more.
• Enterprise AI security challenges.
• How organizations can adopt AI without compromising security.
• Why prompt engineering is becoming a fundamental skill.
• Advice for junior developers entering the AI era.
• The importance of mentoring the next generation of engineers.

• Kilo - AI Developer Platform
• Prompts for Mere Mortals by Florian Hines

One lesson I've learned throughout my career, from enterprise architecture to cloud transformation and now Generative AI, is this.

Every technology revolution follows the same pattern.

At first, companies invest in technology.

Later, they realize they should have invested in people.

I've seen organizations spend millions on software while allocating almost nothing to learning.

Then six months later they ask,

"Why aren't people using the new platform?"

The answer usually isn't difficult.

People don't resist technology.

People resist uncertainty.

When employees understand how AI helps them become better, not replaced, they become curious.

Curiosity leads to experimentation.

Experimentation builds confidence.

Confidence creates innovation.

That's how real transformation happens.

Today's conversation is about something that many organizations overlook.

Not AI models.

Not software.

Not hype.

We're talking about AI capability, the skill that will define the winners of the next decade.

Let's begin.

Episode # 193

John Munsell is the co-founder of Bizzuka, an AI consulting firm focused on artificial intelligence strategy and implementation. With a career spanning over 25 years in marketing, software development, financial services, and sales, John brings a wealth of experience to the AI industry.

• Website: Bizzuka

What Listeners Will Learn:

• Why AI success depends more on people than technology
• The importance of AI skills across an organization
• Why giving employees AI tools isn't enough
• How organizations can measure AI capability
• The hidden cost of poor AI adoption
• Why AI training delivers long-term business value
• Common mistakes leaders make during AI transformation
• Practical ways to improve workforce AI readiness
• Why experimentation matters more than perfection
• How AI can increase productivity across every role

• Bizzuka