Agents After Dark, powered by Prefactor

About Agents After Dark, powered by Prefactor

Agents After Dark is a podcast exploring what actually happens when AI agents move from demos into real-world production environments. Hosted by Matt Doughty (https://www.linkedin.com/in/matthewdoughty/), Co-Founder and CEO at Prefactor (www.prefactor.ai)the show features conversations with founders, enterprise leaders, engineers, and operators building the next generation of agentic systems, AI infrastructure, and runtime platforms. We go beyond the hype to unpack the technical, operational, and organisational challenges behind deploying AI agents at scale — from governance and security to orchestration, observability, MCP, evaluation, and production readiness. If you're building, deploying, or managing AI agents inside modern organisations, this is where the real conversations happen.

Publisher

Matt Doughty

Category

business · technology

Language

en

Explicit

No

First episode

9 Jun 2026

Latest episode

6 Nov 2026

Latest episodes

8 episodes in the feed.

1. 

   6 Nov 2026

   Which SaaS Is Dead? Ask the VC Who Funds It | Josh Farrugia, Principal Investor at Black Nova VC

   Which SaaS is dead? Ask the VC who funds it. Josh Farrugia calls himself a failed founder and a non-technical AI engineer. He has built Black Nova's own AI-native operating system on Claude Code, replacing tools the fund paid tens of thousands of dollars for, while keeping LP data, bank details and personal information well away from it. So the sharper question becomes: Which software should you build yourself, and what should AI never touch? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Josh Farrugia, Principal Investor at Black Nova VC (https://www.linkedin.com/in/joshua-farrugia/), to talk about building real systems without a technical background, and why the best AI builds start with structured data rather than a model. Together they discuss: Why "SaaS is dead" is true for some software and nonsense for the rest, and how to tell which is which How Josh learned to build: hackathons, workshops and borrowing best practice from builders who know more than he does Drawing the line on data: why LP data and bank details stay out of the operating system, and how encryption tools like CipherStash fit in How a team of five core staff and 26 venture partners supports 60 portfolio companies, with fund three adding more Measuring ROI through response times, intro tracking and backlogs cleared Why staff building 15 dashboards to watch their agents is a warning sign, and why enterprise adoption will follow the same path as the cloud Why, if we stopped making new models today, the world would still change overnight How deterministic to be before an agent touches anything, and when that starts to hold better models back Josh's view is that the gap is no longer the technology. It is knowing what AI should do, what it should not touch, and being able to see what it is doing once it is running. This conversation is for founders, investors and operators deciding what to build, what to buy and how far to trust agents with their data. About Josh Josh Farrugia is a Principal Investor at Black Nova VC, an early-stage fund now investing from its third fund. A former founder with a product background, he leads Black Nova's work on AI-native systems. Black Nova is an investor in Prefactor. About Prefactor Prefactor (https://www.prefactor.ai/) makes AI agents better. Prefactor learns how your agents are supposed to behave, evaluates what they actually do on every run, and shows your team how to improve them. Engineers, product managers and the subject matter experts who know what good looks like all work from the same picture, so the agent gets better with every release. Learn more at prefactor.ai (https://www.prefactor.ai/)

2. 

   23 Oct 2026

   What Stops an Agent Paying $20k in $499.99 Chunks? | Hari Om Garg, Managing Consultant at Wipro

   What stops an agent paying a $20,000 invoice in $499.99 chunks? Hari Om Garg designs agentic systems for Fortune 500 companies at Wipro, where an agent might approve any invoice under $500 from a known vendor and send anything bigger to a person. Rules like that only hold if the agent can't learn its way around them, and the target is still a 15 to 25% return on investment in the agent's first year in production. So the sharper question becomes: How do you give an agent real authority without handing it a loophole? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Hari Om Garg, Managing Consultant at Wipro (https://www.linkedin.com/in/hariomgarg/), to break down how he builds guardrails, confidence thresholds and human review into agentic workflows from the start. Together they discuss: Why the $499.99 trick is a fraud detection problem, and how conditional access rules borrowed from cloud sign-in stop it How a confidence wrapper took one agent from 72% to around 90% accuracy, and why anything under 83% goes to a human instead of executing Why every engagement starts with a matrix of quick wins and high-impact use cases, each forecast to return 15 to 25% in its first year Why agents need their own identities, the way cloud platforms already give services managed identities Whether AI ROI is really headcount reduction, and why Hari measures what people do once the agent takes the repetitive work Why new use cases should come before rebuilding legacy RPA bots What a CEO under pressure should do in the next 30 days: cut the noise, spend two weeks finding real use cases, and ship quick wins inside three months As agents move from answering questions to approving payments, the rules around them matter as much as the model. Hari treats guardrails as part of the architecture, designed with the use case rather than added after something goes wrong. This conversation is for CIOs, CTOs and AI leaders who need to show a return on agents this year. About Hari Hari Om Garg is an AI solution architect and managing consultant at Wipro, where he designs agentic AI solutions for Fortune 500 companies. He has previously worked at Microsoft, HCL Technologies, LTIMindtree and PowerObjects, and writes for CIO.com on guardrails and intelligent automation for the enterprise. About Prefactor Prefactor (https://www.prefactor.ai/) makes AI agents better. Prefactor learns how your agents are supposed to behave, evaluates what they actually do on every run, and shows your team how to improve them. Engineers, product managers and the subject matter experts who know what good looks like all work from the same picture, so the agent gets better with every release. Learn more at prefactor.ai (https://www.prefactor.ai/)

3. 

   31 Aug 2026

   Do You Trust the Agent When No One's Watching? | Adam Witanowski, AI Executive

   Only 16% of an engineer's time is spent writing code. That's the number Adam Witanowski keeps coming back to. Hand your engineers the best coding agent on the market and individuals accelerate 120% to 400%. But you've optimised 16% of the pipeline. Measured end to end, he saw between zero and 40% velocity gain, with everything bottlenecking at the PR. So the sharper question becomes: If the coding agents aren't the constraint anymore, what is? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Adam Witanowski, AI architect behind one of Australia's most comprehensive enterprise AI transformations (https://www.linkedin.com/in/witanowski/), to break down what it takes to move an organisation up the agentic maturity curve. Together they discuss: Why rolling out a coding copilot and calling the job done just bloats the middle of your delivery pipeline The harness he built across the full SDLC, and why rework and incidents beat adoption and token spend as metrics every time His audit of 10,000 historical PRs: roughly a third could have been automated, and if you're not ready to automate any of yours, your maturity is lower than you think Governor: policy as code injected into coding agents at invocation time and reasserted on CI, so cyber, legal and compliance shape every build without slowing the dev loop. Human in the loop, not human in the way Skynet: 3,600 repos ingested into a knowledge graph with decision history, so agents see the downstream impact of a change instead of breaking the team next door Why engagement went up while productivity went down, and what job anxiety and a weak front of funnel do to an AI rollout Winning over non-engineers: sit with the call centre, find the pain, and sell teams a solution rather than announcing their replacement His test for where any organisation really sits: maturity isn't about the model. It's about how much you trust the agent when no one's watching it. This conversation is a practical playbook for any engineering or AI leader trying to turn coding agents from an individual productivity toy into an organisational capability. About Adam Adam Witanowski is an AI Executive who spent the past year building one of Australia's most comprehensive enterprise AI transformations from the inside: 11 integrated agentic tools reshaping how engineering teams build software, reimagining the SDLC into an AI-DLC with full executive backing. He shipped 18 products to production in 12 months without writing a meaningful line of code himself. He is now a consulting Chief AI Officer and keynote speaker on AI-powered engineering, with a background in bioinformatics and, formerly, the best job title in the industry: Chief Officer of Creative AI and Novel Engineering. About Prefactor (https://prefactor.tech/) Catch your agents' mistakes before they reach customers. We evaluate every step your agents take, live. Mistakes get caught as they happen. Not in a Slack thread at 1am on a Tuesday. Learn more at prefactor.ai (http://prefactor.ai)

4. 

   11 Aug 2026

   AI Without Seatbelts: Why Every CISO Is Worried About Frontier Models | Sara Abak, CISO, Infra

   Would Toyota release a car without a seatbelt? That's the question Sara Abak keeps coming back to. No carmaker would ship a vehicle without safety features, yet the most powerful software ever built ships with no regulation at all. Every single cyber leader she speaks to is worried about frontier AI models. So the sharper question becomes: How do you secure the enterprise when AI is moving faster than the rules around it? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Sara Abak, CISO across energy and critical infrastructure (https://www.linkedin.com/in/saraabak-255879193), to break down what security leadership actually looks like in the AI era. Together they discuss: Her case for regulating frontier AI like any other product that can cause harm at scale, and why unsafe releases should be paused, not innovation blocked How she turns around security cultures in complex organisations: anchor in what the business actually does, because you can't protect what you can't see Why CISOs get the "department of no" reputation, and how building trust at board level beats laying down the law Why AI features in your ERP or board paper platform stay switched off until they pass a risk assessment Shadow AI in practice: locking down what staff can paste into public AI tools, coaching at the point of upload, and quarterly behaviour reporting so decisions are anchored in data and facts Why standalone responsible AI programs rarely get funded, and why anchoring AI work inside an existing project with legal, HR, and IT at the table is what actually moves things forward How a clinical psychology degree shapes the way she reads the room, wins support, and accepts what she can't change Her three-part answer to frontier AI risk: regulation, layered foundational controls so you're never the easy target, and holding your vendors accountable. And one rule that applies to all of it: never waste a good incident. This conversation is a practical guide to enterprise AI security for any leader trying to enable adoption safely while the technology outruns the rulebook. About Sara Sara Abak is a Chief Information Security Officer with over 20 years in information security and risk, currently CISO & Industry Leader (Energy & Water) at Grampians Wimmera Mallee Water and Founder of Orca One, a fractional CISO advisory. Previously CISO and Chief Privacy Officer at Intellihub, she has secured critical infrastructure across Australia and New Zealand spanning IT and OT, led ISO 27001 certification with zero non-conformances, and was recognised as a Most Influential & Top 100 Cyber Security Leader across ANZ in 2025. She is a certified AI governance leader and ISO/IEC 42001 Lead Implementer. About Prefactor (https://prefactor.tech/) Catch your agents' mistakes before they reach customers. We evaluate every step your agents take, live. Mistakes get caught as they happen. Not in a Slack thread at 1am on a Tuesday. Learn more at prefactor.ai (http://prefactor.ai)

5. 

   18 Jul 2026

   MCP in Prod: Becoming the Electricity Every AI Agent Needs | Ajay Jumar, Head of MCP at Infotrack

   What happens when your customer is an AI agent, not a person? For 25 years InfoTrack has sold authoritative legal and property data to people clicking through interfaces. Now it is making those same services callable by AI agents using the Model Context Protocol (MCP). That shift raises a sharper question: How do you actually run MCP in production inside a regulated enterprise? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Ajay Kumar, Head of MCP Services at InfoTrack (https://www.linkedin.com/in/ajay-kumar87/), to break down what it takes to move Model Context Protocol from experiment to production. Together they discuss: Why InfoTrack stood up a dedicated MCP business unit instead of treating AI as an experiment How an enterprise MCP gateway works in practice: a central registry, a search-and-invoke pattern, and vectorised tools that cut token costs across hundreds of services Why human-in-the-loop elicitation matters the moment an agent can spend money, and why most major providers still don't support it out of the box Why hallucination is an agent problem, not an MCP problem, since MCP itself is deterministic How MCP security is evolving after incidents like the Aura health breach The shift in customer mindset from "what is MCP?" to "make my agent more productive" Why roughly 62% of companies are experimenting with agents while only 28% are scaling, and what separates the two As agents move from assistants to actors that order, transact, and execute, the businesses that win may be the ones whose products are built to be consumed by software. This conversation is a practical guide to running Model Context Protocol in production, covering architecture, security, and go-to-market, for any enterprise starting its own MCP journey. About Ajay Ajay Kumar is Head of MCP Services at InfoTrack, where he leads the company's Model Context Protocol business unit and shapes how AI moves from experimentation into core production workflows. With over 17 years across banking, prop-tech, insurance, and government, Ajay brings an engineering-first perspective on identity, architecture, and enterprise-scale delivery, and on what it actually takes to make agents and MCP work in regulated environments. About Prefactor Prefactor (https://www.prefactor.ai/) helps enterprises trust AI in production. As organisations deploy more AI agents, maintaining visibility into performance, risk, and operational quality becomes increasingly difficult. Prefactor gives engineering, product, and security teams a single platform to monitor AI systems, evaluate outcomes, identify risks, and take action when things go wrong. Learn more at prefactor.ai (http://prefactor.ai)

6. 

   2 Jul 2026

   Trustworthy AI, Where a Mistake Is Life and Death | Aaron Vanston, Co-Founder and CTO of Buildpass

   What happens when a wrong answer isn't a bad response, it's a safety incident? Construction is one of the least forgiving places to deploy AI. The data is scattered across drawings, site diaries, forms, and credentials, and getting it wrong can mean a defect, a blown budget, or an injury on site. BuildPass is putting AI into that environment anyway, without taking the human out of the loop. In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Aaron Vanston, Co-Founder & CTO of BuildPass (https://au.linkedin.com/in/aaronvanston) to unpack how you earn trust in AI agents when the downside is life and death. Together they discuss: Why Aaron went from strategic founder back "on the tools," and how AI lets context-rich, time-poor leaders ship ideas again The case that "SaaS is dead", and why the real opportunity is building the system of record and tools (MCP included) for customers to build on, not another dashboard Why human-in-the-loop is non-negotiable in construction: no hidden multi-step chains, every generative output shown to the user Trust as one-way vs. two-way doors: confidence scores, sources, and impact analysis before an action graduates from "always ask" to autonomous Why delegating control to users beats deciding for them, and how to avoid the "accept all" fatigue that made everyone skip permissions Why evals are a moat: the tests and captured intent-vs-outcome data that tell you whether a new model is actually better for your users How to catch agentic drift when models change underneath you, and why an eval can be as informal as a vibe check A practical guide to running AI agents in a regulated, safety-critical industry, covering human oversight, trust, evals, and quality loops, for any enterprise deploying agents where mistakes actually cost something. About Aaron Aaron Vanston is Co-Founder & CTO of BuildPass, an AI-native construction platform used by hundreds of builders and over 150,000 workers across Australia, New Zealand, and the US. He previously led engineering teams at REA Group and SEEK. About Prefactor Prefactor (https://www.prefactor.ai/) helps enterprises trust AI in production. Prefactor gives engineering, product, and security teams a single platform to monitor AI systems, evaluate outcomes, identify risks, and take action when things go wrong. Learn more at prefactor.ai (http://prefactor.ai)

7. 

   22 Jun 2026

   Writing 90% of Your Code With AI: Vibe Coding at 60 Million Users| Matthew Blode at Linktree

   What happens when AI writes 90% of your code, and 60 million people rely on it? That's the reality for engineers shipping with tools like Claude Code, OpenAI Codex, and Cursor every day. The output is faster and often better, but it raises a harder question: How do you move at that speed without shipping the bug that breaks production? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Matthew Blode, who builds Link Apps at Linktree (https://www.linkedin.com/in/matthewblode/) and is an OpenAI Codex ambassador, to get specific about writing, reviewing, and shipping AI-generated code at scale. Together they discuss: Why Matthew now writes over 90% of his code with AI yet writes less code than ever, and outputs more How vibe coding actually works in production: plan mode, phased to-dos, and layered review across multiple coding agents plus human code review The invoicing bug that reached production when an AI hallucinated a dependency upgrade, and the QA that would have caught it How risk tolerance shifts from a startup MVP to a platform with 60 million users, using feature flags, internal dogfooding, and AI code reviewers in GitHub Why AI evals still need humans in the loop ("who watches the watchman"), and the observability tooling Linktree leans on Delivering AI features users actually want, including the right to opt out of AI entirely, and the data sovereignty question that raises The Pixar storyboard method for moving from proof of concept to production without burning money on dead-end demos Whether you are a founder shipping an AI-native MVP or an engineer delivering features to millions, this is a practical look at agentic coding in production: how to keep velocity high and quality intact when the machine writes most of the code. About Matthew Matthew Blode builds Link Apps at Linktree and is an OpenAI Codex ambassador, deep in the AI coding and Codex community. He has built and sold two startups, including Fingertip, an AI-powered website builder, and brings a hands-on perspective on agentic coding, code quality, and shipping AI features safely at scale. About Prefactor Prefactor (https://www.prefactor.ai/) helps enterprises trust AI in production. As organisations deploy more AI agents, maintaining visibility into performance, risk, and operational quality becomes increasingly difficult. Prefactor gives engineering, product, and security teams a single platform to monitor AI systems, evaluate outcomes, identify risks, and take action when things go wrong. Learn more at prefactor.ai (http://prefactor.ai)

8. 

   9 Jun 2026

   What Does an AI Agent Buy? | Andrew Bird, Head of AI at Affinda

   What happens when software becomes the customer? For decades, products have been designed, marketed, and sold to humans. But as AI agents become increasingly capable of researching, evaluating, and making decisions on our behalf, a new question emerges: What does an AI agent actually buy? In this episode of Agents After Dark, Matt Doughty (https://www.linkedin.com/in/matthewdoughty/) sits down with Andrew Bird, (https://www.linkedin.com/in/andrew-bird/) Head of AI at Affinda, to explore how AI is changing the way products are discovered, evaluated, and purchased. Together they discuss: What happens when agents become decision-makers rather than assistants How software discovery changes in an agent-driven world Why traditional go-to-market motions may need to evolve How trust, reputation, and verification influence agent decisions The future of procurement in a world of autonomous systems What founders and product teams should be building for as AI adoption accelerates As agents move from copilots to autonomous actors, the way products are bought, sold, and evaluated may fundamentally change. This conversation explores what that future could look like and what it means for businesses today. About Andrew Andrew Bird is Head of AI at Affinda, where he helps organisations apply AI to automate and extract value from complex, unstructured information. Having worked at the forefront of AI adoption through multiple technology waves, Andrew brings a practical perspective on how agents, automation, and AI-driven decision making are changing the way software is built, evaluated, and purchased. About Prefactor Prefactor (https://www.prefactor.ai)helps enterprises trust AI in production. As organisations deploy more AI agents, maintaining visibility into performance, risk, and operational quality becomes increasingly difficult. Prefactor gives engineering, product, and security teams a single platform to monitor AI systems, evaluate outcomes, identify risks, and take action when things go wrong. Learn more at prefactor.ai (http://prefactor.ai)

Reach and audience

Public platform figures. Ratings count people who left a rating, not total listeners.

Spotify

5.0 / 5

3 ratings

Contact Agents After Dark, powered by Prefactor

Guest appearances

Books guests

Based on episode analysis; this does not confirm that the show is currently accepting guests.