Winners' Circle

Harshit Kohli is helping enterprises understand where AI, cloud, streaming analytics, and cybersecurity are heading next. As a Senior Technical Account Manager at Amazon Web Services and a GenAI streaming specialist, Harshit works with customers on cloud adoption, artificial intelligence adoption, and real-time data use cases. He is also pursuing a doctorate in artificial intelligence and continues to serve as one of BIG’s All-Star Judges.

In this episode, Russ and Harshit discuss what he is seeing across AI, cybersecurity, MCP, and agentic systems after speaking at industry events including RBLN and the MCP Dev Summit. Harshit explains why companies are moving quickly toward AI, but must also keep security, governance, authorization, and authentication at the center of every deployment.

The conversation dives into model context protocol, or MCP, and how it is evolving beyond simple request and response workflows. Harshit shares how he demonstrated a streaming MCP architecture using Amazon MSK, WebSockets, IAM authorization, and Amazon Bedrock to push real-time context into AI agents. The goal is to help organizations detect anomalies, understand root causes, and act faster when minutes or seconds can carry major business risk.

Russ and Harshit also explore the risks around compromised MCP servers, agent hijacking, shadow AI, least privilege access, sandbox execution, human in the loop validation, and the growing need for governance around AI agents that can now plan, act, execute code, and manage infrastructure.

Along the way, Harshit shares why enterprises should not chase AI for its own sake, why many projects may fail without clear use cases and controls, and why the best companies are not replacing people with AI, but investing in their teams so people can do more with it.

Topics Covered:

[00:00] Welcome and intro, Harshit Kohli and BIG’s All-Star Judges

[00:43] Harshit’s role at AWS and work in GenAI streaming

[01:55] His doctorate in artificial intelligence and industry speaking

[02:19] Key themes from RBLN around AI, cybersecurity, and governance

[03:56] AI-driven phishing volume and security industry response

[04:33] Why governance and security must be built into AI products

[05:44] MCP Dev Summit and the evolution of model context protocol

[06:23] Moving MCP beyond basic request and response

[07:00] Harshit’s streaming MCP demo with Amazon MSK, WebSockets, IAM, and Bedrock

[08:26] Why real-time anomaly detection matters in financial systems

[08:53] Root cause analysis and remediation through AI agents

[10:29] Observability, telemetry, and cost anomaly detection

[12:16] MCP design flaws, exposed servers, and enterprise risk

[12:52] Why AI agents now execute code and manage workflows

[14:41] What a compromised MCP server means

[15:05] Sandbox execution and least privilege access

[17:23] Consent, verification, and human in the loop controls

[18:40] The rise of enterprise AI agents

[19:21] Why agent autonomy creates governance challenges

[20:38] Deploying agents versus deploying applications

[22:38] Agent hijacking and prompt injection risks

[23:00] Poisoned documents, malicious tools, and compromised MCP servers

[25:19] Patch windows, enterprise readiness, and AI cyber risk

[26:30] Shadow AI inside organizations

[26:55] Employees using unapproved AI tools at work

[28:57] Data leakage, compliance violations, and hallucination risk

[30:46] Why some AI projects may be canceled

[31:07] Starting with real use cases instead of forcing AI

[32:44] Whether AI has become infrastructure yet

[33:39] What the best companies are quietly getting right

[34:46] Final thoughts on investing in people while adopting AI

Sasibhushan Rao Chanthati, Senior Software Engineer at Hirekeyz and a two time BIG All Star Judge for Information Technology, joins the Winners' Circle to talk about agentic AI, the FinOps boom, small versus large language models, and his own original research into detecting IT burnout with AI.

Guest info: Sasibhushan Rao Chanthati, Senior Software Engineer, Hirekeyz.

• [00:00] Welcome and congratulations on being a two time BIG All Star Judge
• [01:02] Sasi's role at Hirekeyz and his previous work at T. Rowe Price
• [02:55] What drew Sasi to judging IT nominations in the first place
• [04:41] Task specific agents versus chatbots, the real difference
• [10:13] Small language models versus large frontier models
• [13:25] The hybrid model trend engineers are actually using
• [14:58] Why cheaper AI products often cut corners on security
• [18:23] Customer support automation and where it is heading
• [22:08] The biggest shifts Sasi has seen in the last two years
• [25:12] Why FinOps is becoming mission critical
• [27:00] The origin story behind Sasi's AI burnout detection research
• [29:04] The moment that convinced Sasi this was worth building
• [32:27] Is AI making burnout better or worse
• [37:39] Closing thoughts and congratulations

• After two years judging IT nominations for BIG, Globee, and Stevie, Sasi has learned that the strongest submissions describe a specific mechanism, what was actually built and how it gets used, rather than leaning on company size or mission statements alone.
• AI cost management went from a niche concern to a near universal practice in just two years, with FinOps Foundation research showing adoption jump from 31 percent of organizations in 2024 to 98 percent in 2026, as AI spend joins cloud spend as something every technology leader has to actively manage.
• Sasi's own research into AI driven IT burnout detection, built using vector embeddings and workplace communication analysis, is one of the few systematic attempts to measure a problem the industry has talked about informally for years but rarely tried to quantify.

• Hirekeyz
• Sasibhushan Rao Chanthati on LinkedIn
• Sasi's research on Google Scholar
• Sasi's research on ResearchGate
• Sasi's ORCID

Sasi is a two time BIG All Star Judge for Information Technology. See his full judge profile.

🔗 Subscribe to the Winners' Circle podcast.

Amy Worley is helping business leaders rethink privacy, cybersecurity, AI governance, and data protection as one connected challenge. As a Managing Director at BRG, Amy brings a rare mix of experience as a former trial lawyer, in-house privacy leader, consultant, expert witness, and author of The Confidence Advantage. BRG recently won an AI Excellence Award, recognizing work at the intersection of AI, privacy, cybersecurity, and governance.

In this episode, Russ and Amy explore why AI has made it harder for companies to treat privacy, security, and governance as separate functions. Amy explains how businesses often have legal teams talking about GDPR or HIPAA, cybersecurity teams talking about threat actors and attack surfaces, and AI governance teams working in still another language. Her Confidence by Design framework brings those worlds together through a shared set of principles, common language, and unified risk metrics.

Amy also shares how her career shaped her perspective. She began in law, moved through data privacy and breach response as the internet and privacy statutes evolved, then went in-house to build a GDPR program for a multinational pharmaceutical company. That experience taught her the gap between giving advice and actually building programs that work inside a business.

The conversation also covers what happens during data incidents, why communication and decision authority often break before technical response does, and how Amy uses a “pre-mortem” process to help companies identify what could derail a governance program before it starts.

Russ and Amy also discuss AI deployment, data debt, enterprise LLMs, accountability, board responsibility, chief trust officers, and why digital trust should not be treated as a cost center. Amy’s message is clear: in a digital first, AI powered world, evidence based trust can become a real business advantage.

Topics Covered:

[00:00] Welcome and intro, Amy Worley, BRG, and the AI Excellence Award

[00:22] What BRG does as a multinational expert services firm

[00:53] Amy’s background as a lawyer and privacy professional

[01:11] What a former trial lawyer sees in data breach response

[02:00] Moving from legal advice to building real privacy programs

[03:13] The Confidence Advantage and Confidence by Design

[03:33] Why privacy, cybersecurity, and AI governance need to be unified

[04:00] Building an 11 principle framework across three disciplines

[05:05] What breaks when privacy, security, and AI teams are siloed

[06:00] Creating a common language for executives and risk

[07:05] Whether one team should own the unified governance vision

[07:59] What day one looks like in a data incident or program build

[08:14] Communication rules and decision authority during incidents

[09:00] Using a pre-mortem to identify why a program might fail

[10:07] Common roadblocks: executive understanding and team bandwidth

[10:45] Defining what winning looks like before the work begins

[12:08] What courtroom experience teaches about documenting governance

[13:28] How AI responsibility has shifted from planning to cleanup

[14:15] Why companies now need diagnostics for AI bottlenecks

[15:00] Building agile governance and risk tiers for AI adoption

[16:03] Confidence by Design in thirty seconds

[16:40] Maximizing the value of business data

[17:32] Data debt, enterprise LLMs, and old information resurfacing

[19:49] How to identify who is truly accountable for risk

[20:09] Why AI governance is becoming a board level issue

[21:20] The case for a chief trust officer

[22:18] What leaders should do differently tomorrow

[22:31] Digital trust as a competitive advantage

[23:22] Final thoughts on AI, privacy, cybersecurity, and the future of trust

Matt Spiegel is helping law firms rethink growth, client relationships, and the business side of practicing law. As CEO and Co-Founder of Lawmatics, Matt leads a legal tech company built to help law firms manage leads, automate marketing, track performance, improve intake, and strengthen the client journey from first contact through future referrals. Lawmatics recently won a Best Places to Work award, recognizing the culture Matt and his team have built as the company continues to grow.

In this episode, Russ and Matt discuss how Lawmatics became much more than a CRM for lawyers. Matt describes it as a growth platform for law firms, supporting lead management, marketing automation, reporting, analytics, and the relationship moments that happen before and after a client matter.

Matt also shares the story that led him into legal tech. As a criminal defense attorney, he received a bar complaint from a client that came down to one issue: communication. That experience helped him see how common slow follow-up and poor client communication were in the legal industry, which led him to launch MyCase and later return to the market with Lawmatics.

The conversation also explores Lawmatics’ culture and why in-person collaboration matters to Matt. The company works in office Monday through Wednesday, which Matt believes creates the spontaneous conversations, customer insights, and team connections that are hard to replicate remotely.

Russ and Matt also discuss the changing legal market, the rise of competitive legal marketing, Lawmatics’ expansion into personal injury, and how the company is building AI into a broader platform rather than relying on AI as a standalone feature. Matt also shares advice for founders building niche B2B SaaS companies and for lawyers who want to grow their firms like modern businesses.

Topics Covered:

[00:01] Welcome and intro, Matt Spiegel, Lawmatics, and the Best Places to Work award

[00:31] What Lawmatics does beyond CRM

[01:11] Matt’s background as a criminal defense lawyer

[01:29] The client complaint that sparked his legal tech journey

[02:00] Starting MyCase and building a practice management platform

[02:38] Why Matt returned to legal tech with Lawmatics

[03:45] Lawmatics’ culture and Best Places to Work recognition

[04:02] Scaling culture as the company grows

[05:00] Why Lawmatics is in office Monday through Wednesday

[06:20] How in-person work sparks customer and product insights

[07:29] Why random conversations can create great ideas

[08:49] Building a company around smart, interesting people

[09:44] Getting lawyers to adopt new technology

[10:22] Why the market was not ready for Lawmatics earlier

[11:00] How legal marketing changed law firm operations

[12:10] Why consumer-focused law firms are more competitive

[12:42] Personal injury, MSOs, ABSs, and new legal market dynamics

[14:01] Lawmatics’ UI overhaul and AI-assisted features

[14:26] How customer problems guide product development

[15:19] Expanding deeper into personal injury workflows

[16:00] Building an AI suite on top of a broader platform

[17:01] Growth opportunities across practice areas and firm types

[18:12] Legal marketing, pay-per-click, and client acquisition

[19:06] Matt’s advice for founders building niche B2B SaaS products

[20:00] Why AI alone is not enough to build a durable company

[20:14] Matt’s advice for lawyers growing a law firm

[21:23] Final thoughts on Lawmatics, culture, and legal tech growth

Chris Shyrock is helping Atlantic.Net strengthen customer support in one of the most demanding environments in technology: cloud infrastructure and data centers. As Director of Support Services, Chris leads the team responsible for supporting customers who rely on Atlantic.Net for cloud services, colocation, dedicated hosting, managed services, and infrastructure they cannot afford to have fail.

In this episode, Russ and Chris explore how Atlantic.Net transformed its support organization while also completing a major data center migration with no customer impacts attributed to the move. Chris explains how the company improved queue ownership, daily governance, shift handoffs, escalation paths, and customer communication while maintaining 24/7/365 service for clients in high-stakes industries like healthcare.

They also discuss what it takes to turn technically strong support staff into customer-ready communicators. Chris shares how he reviewed calls and emails, re-onboarded the team, trained people to own the call, and encouraged support professionals to de-escalate concerns, explain issues clearly, and meet customers at their level.

Along the way, Chris discusses Trustpilot reviews, customer feedback, team empowerment, regulated customer environments, AI cloud infrastructure, NVIDIA GPU hosting, continuous improvement, and why support is often the trust layer between a technology provider and the customers who depend on it.

Topics Covered:

[00:01] Welcome and intro, Chris Shyrock, Atlantic.Net, and the Excellence in Customer Service Award win

[00:45] Atlantic.Net’s long history in cloud and infrastructure services

[01:23] What Atlantic.Net provides for cloud, colocation, dedicated services, and managed infrastructure

[02:58] Why Atlantic.Net focused on customer service transformation

[04:00] Teaching technical support teams to communicate with customers more clearly

[05:21] Aligning staffing, mindset, and shift coverage

[06:30] Queue ownership and daily governance in support operations

[07:26] Managing customer change, technology change, and internal change at the same time

[08:45] Training the team to own the call and de-escalate urgent customer issues

[09:25] Why practice matters in customer support training

[11:00] Getting support teams comfortable with doing the harder thing for customers

[12:10] Why leaders need to be reachable during high-stakes incidents

[13:40] Transforming the support team during a major data center migration

[14:46] How the support organization contributed to a no-impact migration

[16:10] How major operational events can elevate a team

[16:44] The hardest habit to change in a long-running support organization

[17:44] Using Trustpilot reviews as evidence of support quality

[19:34] Why moving review scores is hard and meaningful

[20:00] Turning customer problems into trust-building opportunities

[22:36] Upskilling support teams for new infrastructure and AI workloads

[23:32] Advice for technical support leaders with inconsistent customer experience

[27:22] Final thoughts on Atlantic.Net’s transformation and continuous improvement

Kim Canning is helping Kaplan expand access to career readiness, test prep, licensure prep, credentialing, and professional development resources through a new model built for universities and students. As VP of Strategic Partnerships at Kaplan, Kim works with colleges and universities to bring Kaplan’s All Access License to learners at scale. Kaplan recently won a BIG Awards for Business honor for the program.

In this episode, Russ and Kim explore how Kaplan has evolved over more than 80 years from test prep in Stanley Kaplan’s Brooklyn brownstone to a modern education company supporting students, universities, and career pathways across the country.

They dive into Kaplan’s All Access model, which gives universities an enterprise license so students can access a broad catalog of Kaplan resources without each student having to pay individually. Kim explains how the program helps remove financial barriers while also introducing students to resources they may not have known existed, from MCAT and LSAT prep to critical thinking, data literacy, career readiness, credentialing, and licensure support.

The conversation also covers how universities are using All Access in different ways, including first-year programs, career preparation, graduate school pathways, healthcare licensure, and support for historically black colleges and universities. Kim shares examples from Hampton University, NYU, Illinois public institutions, and other partners using the program to support retention, persistence, graduation, and student confidence.

Along the way, Kim discusses education equity, soft skills, COVID-era learning gaps, critical thinking, healthcare pipelines, first-generation college students, government partnerships, and why universities may need to think differently about the resources students need to succeed.

Topics Covered:

[00:22] Welcome and intro, Kim Canning, Kaplan, and the BIG Awards for Business win

[00:49] Kim’s 28-year journey with Kaplan

[02:12] Kaplan’s 80-plus year history and Stanley Kaplan’s founding mission

[03:25] What the Kaplan All Access License is

[03:44] Moving from department-level purchases to university-wide access

[06:07] How universities use All Access across different student needs

[07:23] Cost, awareness, and access barriers for students

[08:00] Why students may not know which resources can open doors

[09:26] Helping students reduce anxiety and understand expectations

[09:55] How All Access can support engagement, retention, and loyalty

[11:37] Flipping the traditional student-pays model

[12:40] Kaplan’s footprint with historically black colleges and universities

[14:00] Why All Access can create immediate student impact

[15:02] Supporting freshmen and sophomores with foundational skills

[15:42] Hampton University and critical thinking as a student success priority

[17:40] Why soft skills matter more in an AI-driven world

[18:49] Building critical thinking resources beyond test prep

[21:30] Supporting science courses and the medical school pipeline

[22:43] The Illinois All Access initiative and statewide education equity

[24:00] Supporting 17 Illinois institutions and saving students millions

[24:43] How All Access can help first-generation and underrepresented students

[25:27] Student and advisor feedback from the Illinois rollout

[26:47] Coordinating across government, institutions, and education leaders

[27:25] Why mission alignment helped move the Illinois initiative forward

[30:06] Why All Access is one piece of a larger higher education puzzle

[31:00] Personalizing student experiences and supporting future pathways

[31:57] Why All Access complements universities rather than replacing them

[33:24] Final thoughts on improving the student experience and expanding access

Stephen Garcia is helping BreachRX bring structure, governance, and accountability to one of the most chaotic moments any organization can face: a cyber incident. As Chief Information Security Officer at BreachRX, Stephen brings years of experience from the customer side of cybersecurity, including work across financial services, technology, gaming, and enterprise environments. BreachRX recently won a Fortress Cybersecurity Award for its cyber incident response management platform.

In this episode, Russ and Stephen explore why most organizations have tools to detect and contain incidents, but often lack a governance layer for everything that happens once a technical event becomes an enterprise crisis. Stephen explains how BreachRX helps coordinate legal exposure, regulatory deadlines, communications obligations, executive accountability, and audit trails in real time.

They dive into the new category of cyber incident response management, or CIRM, and why binders, conference calls, and manual call trees are no longer enough. Stephen shares what the first hours of a serious incident can look like from the inside, why chaos often emerges even when plans exist, and how organizations can prepare more effectively before the crisis begins.

The conversation also covers RexAI, BreachRX’s generative AI engine built specifically for incident response, as well as Mobile Command, out-of-band communications, regulatory readiness, tabletop exercises, executive liability, and the company’s CIRM warranty.

Along the way, Stephen discusses resilience, trust, accountability, legal timing, simultaneous incidents, enterprise risk, AI agents, board communication, and why the organizations that survive the next wave will be the ones that can compress the time between knowing and doing.

Topics Covered:

[00:01] Welcome and intro, Stephen Garcia and BreachRX’s Fortress Cybersecurity Award win

[01:03] What BreachRX does and why cyber incidents need a governance layer

[01:30] Moving from technical containment to enterprise crisis management

[01:57] Why Stephen switched from the customer side to BreachRX

[02:17] The importance of managing incident chaos

[03:54] What the first four hours of a serious incident can look like

[04:14] Why preparation, logging, and lessons learned matter

[07:05] Why response plans often fall apart under pressure

[07:59] Handling multiple simultaneous incident inputs

[09:29] BreachRX as a coordination layer and incident response fabric

[09:46] Out-of-band communications during ransomware and major disruptions

[10:30] Pulling in the right teams, including legal, at the right time

[11:00] Why slowing down can help organizations speed up

[12:52] Building governance structures before a crisis begins

[14:19] What convinced Stephen that the BreachRX platform worked

[15:20] Regulations, legal workflows, and global response requirements

[16:42] Using tabletop exercise budgets to bring BreachRX into an organization

[17:49] Why gaps and leaks can kill incident response

[20:34] Why BreachRX’s warranty turns software into a trust decision

[21:04] RexAI and purpose-built generative AI for incident response

[21:35] How RexAI guides responders in high-pressure environments

[22:18] Mobile Command and managing incidents from anywhere

[24:03] Compressing the time between knowing and doing

[24:22] How AI changes the incident response landscape

[25:18] Expanding the definition of an incident beyond major breaches

[25:40] Why IT, security, and business risk are increasingly connected

[27:20] Security as trust management

[28:00] What CEOs and boards should understand before the next breach

[28:58] Final thoughts on BreachRX, response coordination, and cyber resilience

Avi Kedmi is helping SysAid rethink what customer service can look like in the age of AI. As CEO and Founder of SysAid, Avi leads a global IT service management company serving mid-sized clients in more than 100 countries. SysAid recently won an Excellence in Customer Service Award for its transformation of customer care using its own AI-driven tools, systems, and processes.

In this episode, Russ and Avi explore how SysAid improved CSAT from 68 to 93, reduced detractors by 94 percent, cut escalations, reduced P1 incidents, shortened call times, and increased frontline resolution rates. Avi credits Asaf Goldstein and the care and support team for leading much of the transformation.

They dive into what it means to become an agentic organization. Avi explains why companies need more than chatbots to make AI work. They need a strong data foundation, broad API access across systems, governance, monitoring, guardrails, and smart people in the middle who can identify and deploy high-value agentic use cases.

The conversation also covers how SysAid uses AI to detect patterns, summarize issues, route tickets, identify unhappy customers, alert customer success teams, and help support teams act faster. Avi shares why the company is focused on using AI for work that is “unhuman,” while making human interactions more personal, compassionate, and meaningful when they matter most.

Along the way, Avi discusses AI adoption, change management, employee training, customer empathy, proactive support, the future of IT service management, and why AI should help people do more, not make them feel replaced.

Topics Covered:

[00:01] Welcome and intro, Avi Kedmi and SysAid’s Excellence in Customer Service Award win

[00:35] What SysAid does in IT service management

[01:21] AI, uncertainty, and the biggest priorities for SysAid’s leadership team

[02:00] How AI is redefining what can be built and how fast

[02:45] Product AI, customer adoption, and building trust around AI

[03:20] What it means to transform into an agentic organization

[03:41] Helping employees adopt AI while feeling valued

[04:55] SysAid’s customer service transformation results

[06:07] Giving credit to Asaf Goldstein and the support leadership team

[06:45] The two foundations of an agentic organization

[07:10] Why the data lake strategy has changed in the AI era

[07:50] Why systems need broad API and MCP access

[09:46] How SysAid prioritized which AI use cases to tackle first

[10:08] Using correlation to understand what drives detractors and poor CSAT

[11:26] Why some support tasks are “unhuman” and ideal for AI

[12:33] Why SysAid chose agentic AI instead of only building chatbots

[13:08] The difference between chasing chatbots and controlling the AI beast

[13:45] AI certainty thresholds and fast human failover

[14:42] Connecting support, customer success, R&D, leadership, and product

[15:02] Breaking silos with AI across the company

[16:15] How support teams adapted to new AI tools and workflows

[17:39] Why fast human response matters when AI falls short

[18:26] Making AI-supported service feel more personal and compassionate

[19:02] Adding agent names, photos, and human identity into customer interactions

[20:41] How SysAid’s culture changed through the AI transformation

[20:52] The pace, pressure, and opportunity of becoming an AI-first organization

[22:15] What Avi learned about change management

[23:49] The future of predictive and preventative support

[24:11] How AI could resolve bugs and tickets end to end

[26:18] How customer service may look three years from now

[27:00] Why there may be fewer tickets in the future

[29:37] Final thoughts on SysAid’s transformation and award recognition

Dr. Tina Srivastava is helping Badge rethink authentication for a world where humans, machines, and AI agents all need to prove they are who they say they are. As Co-Founder of Badge Inc., Dr. Tina is working to eliminate stored secrets from authentication and create a more secure root of trust for identity. Badge recently won a Fortress Cybersecurity Award for its work building a trust layer for the AI era.

In this episode, Russ and Dr. Tina explore why traditional authentication has become so frustrating and vulnerable. Passwords, passkeys, push notifications, knowledge-based questions, stored biometric data, and static credentials all create risks, especially as AI makes phishing and impersonation attacks more convincing.

They dive into Badge’s secretless authentication model and how the company allows users to derive a cryptographic key on the fly without storing private data in a central database. Dr. Tina explains how her background in national security and the Office of Personnel Management breach, which exposed millions of fingerprints, shaped the belief that people should not have to give up private data just to prove who they are.

The conversation also covers fuzzy extraction, biometric privacy, zero standing privileges, workload authentication, and the growing need to authenticate AI agents safely. Dr. Tina explains how Badge can help ensure that agents only receive the permissions they need, only when they need them, and only when the verified human authorizes the task.

Along the way, Dr. Tina discusses phishing-resistant authentication, agentic AI, identity without secrets, OEM partnerships, Badge as an embedded trust layer, and why authentication must become more seamless, private, and secure if people and companies are going to fully trust AI-powered systems.

Topics Covered:

[00:00] Welcome and intro, Dr. Tina Srivastava, Badge Inc., and the Fortress Cybersecurity Award win

[00:25] Why authentication has become more complicated and more important

[00:50] How Badge turns authentication on its head

[01:31] The founding insight behind Badge and identity without secrets

[01:48] The OPM breach and why stored biometric data creates lasting risk

[02:43] Zero knowledge authentication and proving identity without stored secrets

[03:24] Deriving keys on the fly and making the user the root of trust

[04:28] What the user experience looks like without passwords

[04:42] Supporting face, fingerprint, context, device characteristics, and PIN factors

[05:20] Multi-factor authentication for workloads and removing static API credentials

[05:50] Why agentic AI creates new permission and access challenges

[06:37] What biometric fuzzy extraction means for privacy

[07:33] Using fuzzy inputs to derive precise cryptographic keys

[08:00] How agentic AI changes authentication and attack surfaces

[08:31] Why AI makes phishing attacks more personalized and harder to detect

[09:30] Using Badge to verify human approval before agents take sensitive actions

[10:50] Zero standing privileges for AI agents

[11:20] Giving agents temporary permissions only when authorized

[12:13] Governance layers for employees and AI systems

[13:08] Badge as an embedded “Intel inside” authentication layer

[13:49] Working with existing identity providers and enterprise systems

[14:12] Badge’s partner ecosystem and complementary integrations

[15:31] Growth through OEM and embedded trust relationships

[16:17] Authenticating humans, machines, and AI agents

[16:55] Why AI agents should be treated like new employees that need guardrails

[17:37] What happens if trust erodes in AI systems

[18:20] Making secure access seamless across devices, systems, and environments

[19:28] Final thoughts on authentication, quantum complexity, and the future of trust

Theresa Bui is helping SymphonyAI bring operational AI into some of the world’s largest enterprises. As CMO of SymphonyAI, Theresa works across a company serving customers in retail, financial services, manufacturing, enterprise IT, and media. SymphonyAI recently won three AI Excellence Awards for products built to help companies deploy AI at scale.

In this episode, Russ and Theresa explore what makes SymphonyAI a vertical AI company and why that matters for enterprise adoption. Theresa explains how the company enters industries with prebuilt agents, ontologies, and models that already understand industry-specific workflows, allowing customers to move from implementation to use cases much faster than traditional horizontal AI platforms.

They dive into Eureka, SymphonyAI’s foundational platform for enterprise AI. Theresa walks through its three core layers: shared industry context through domain knowledge graphs, adaptive orchestration that applies the right AI tool to the right workflow, and governance from day one so every decision point can be logged, reviewed, and trusted.

The conversation also covers how Eureka works in real-world environments, from industrial manufacturing and predictive asset management to financial services investigations. Theresa shares examples of agentic AI detecting a worn part, checking inventory, creating a work order, and scheduling maintenance within minutes, as well as helping banks reduce investigation time while keeping humans in the loop.

Along the way, Theresa discusses AI sovereignty, data ownership, human oversight, regulatory trust, scaling beyond pilots, and why the hardest part of enterprise AI is often not proving it works once, but making it work across hundreds of plants, workflows, and business units.

Topics Covered:

[00:01] Welcome and intro, Theresa Bui, SymphonyAI, and the AI Excellence Awards wins

[01:23] What SymphonyAI does as a pure play AI company

[02:00] Why SymphonyAI defines itself as a vertical AI company

[02:30] Horizontal AI versus vertical AI in enterprise deployments

[03:00] Prebuilt agents, ontologies, and models for industry-specific use cases

[03:46] How Eureka supports enterprise AI deployment

[04:20] Eureka’s three foundations: context, orchestration, and governance

[04:40] Domain knowledge graphs and shared industry context

[05:25] Adaptive orchestration and using the right AI tool for each workflow

[06:00] Governance, audit trails, and trust in regulated industries

[07:20] How Eureka powers industrial manufacturing use cases

[08:00] Predictive asset management, process optimization, and frontline worker workflows

[08:25] Agentic AI example: worn part detection, inventory checks, work orders, and maintenance scheduling

[10:08] Perceive, reason, act and giving AI a structured path to solve problems

[10:48] Human in the loop controls and tolerance thresholds

[11:20] SymphonyAI Risk Intelligence for financial services investigations

[12:00] How banks can adjust oversight as AI earns trust

[13:14] How AI learns from human overrides and decision nuance

[14:01] What AI sovereignty means in enterprise environments

[15:00] Data ownership, private tenants, hosted environments, and competitive advantage

[16:35] Why sovereign AI can become a company’s proprietary IP

[18:05] What enterprise leaders misunderstand about AI implementation

[18:36] Why scaling AI is harder than running a successful pilot

[19:20] The challenge of deploying across many plants, lines, and use cases

[20:00] Why data normalization and shared ontologies matter for scale

[20:48] Comparing AI deployment to outsourcing processes and documenting workflows

[21:30] Final thoughts on SymphonyAI’s growth, awards, and enterprise impact