AWS for Software Companies Podcast

Omilia’s CTO shares their strategy on building AI that facilitates billions of calls and what it takes to win the next decade.

Topics Include:

• Miguel Alava welcomes Marios Fakiolas, CTO of Omilia
• Omilia has built production AI for over 20 years
• Banking and telco clients demand speed and accuracy
• Omilia builds its own agent framework and self-learning agents
• Infrastructure and data matter more than any single model
• Models are ships; Omilia's infrastructure is the permanent dock
• AI is core infrastructure at Omilia, not an external API
• Builders differ from orchestrators by owning bespoke models
• Platform is a kitchen; models are ingredients for recipes
• Omilia believes AI should be accessible, not just for elites
• AI vendors split into camps by economics and scalability
• Gen AI and ROI don't yet align well industry-wide
• Small unaddressed pain points can quietly sink AI projects
• Omilia revisits its offering using deep customer knowledge
• Cost-efficient economics at billions of calls is Omilia's moat
• Making AI work differs from making AI profitable
• Bedrock enables fast prototyping and early customer feedback
• Omilia moves to SageMaker AI to fully own its models
• Marios praises the AWS team supporting Omilia daily
• Speed round covers AI advocates, cloud, and adaptability ahead

Participants:

• Marios Fakiolas – Chief Technical Officer, Omilia
• Miguel Alava – EMEA ISV General Manager, Amazon Web Services  

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

From alert to root cause in one minute - how PagerDuty built autonomous incident response on Amazon Bedrock, and the future of triage and trust.

Topics Include:

• PagerDuty's agents must perform during 2am outages — stakes are high
• Software shipping accelerated dramatically; production environments largely did not
• A 9:30pm slowdown traced to a race condition solved two years earlier
• The fix was documented — but the context wasn't at hand
• PagerDuty Advance ships four agents: SRE, Scribe, Shift, Insights
• Why four, not one? Focus and predictability in non-deterministic systems
• Saurabh Shanbhag: Bedrock is far more than a model service
• Zero data retention, PrivateLink, TLS — why enterprises pick Bedrock
• Frontier models everywhere burns tokens; classify, route, distill, fine-tune
• SRE agent triages alerts before you even join the call
• One minute to root cause — context beat raw intelligence
• Human surfaces versus machine surfaces: MCP and CLI move fastest
• "The model eats the harness" — every upgrade invalidates foundational components
• Feeding agents everything failed; compartmentalised investigation threads work better
• New York Life's three stages of trust, and the seatbelt override that wasn't

Participants:

• Tom Hogarty - Senior Director Product Management, PagerDuty
• Saurabh Shanbhag – Sr Partner Solution Architect, Amazon Web Services  

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

Domo and AWS reveal how AI agents freed sales reps from 20 hours of weekly busywork, turning scattered data into real-time coaching and forecasting.

Topics Include:

• Domo and AWS teams introduce today's session on AI agents in sales.
• Topic: using AI agents to transform sales operations, from insight to action.
• IT teams increasingly asked to turn data into actionable outcomes, not just access.
• Domo's CRO wanted AI agents to boost sales rep efficiency significantly.
• Reps act like "archaeologists," digging through scattered systems for basic context.
• This digging eats roughly 20 hours weekly, half of reps' time.
• Goal: personal AI agent per rep, understanding their book of business.
• Live demo begins: agent app surfaces urgent items needing attention.
• Agent tracks deal milestones, timelines, and forecasts from call and email data.
• "Deal coach" feature grades rep performance and suggests next actions.
• Agent tone can be tuned from gentle to direct, aiding tough feedback.
• Architecture overview begins: building an AI-ready data foundation first.
• Data from CRM, calls, and emails flows into a cloud warehouse.
• Two agents built: automated deal analysis and personalized deal coach.
• Agents write insights back to CRM, preserving human edit control.
• Recipe: build foundation, activate with agents, distribute to people.
• Governance must be embedded throughout, not bolted on afterward.
• Second example: Fogo do Chão uses AI to analyze restaurant reviews.
• AWS architecture explained: Domo runs on Bedrock, defaulting to Anthropic models.
• Q&A: sales team adoption was immediate and enthusiastic post-rollout.

Participants:

• Jason Longhurst – Head of Product Marketing, Domo
• Aman Tiwari - Sr Solutions Architect, ISV, Amazon Web Services  

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

Curious how AI can query your enterprise data without moving it or making things up? AWS and Teradata break down a trustworthy analyst agent built for real production use.

Topics Include:

• Neha Wadhera (AWS) introduces Trinath Yarlagadda and the Teradata Analyst Agent
• Enterprise AI data prep is costly, stalling most orgs at experimentation
• Agent answers plain-English questions via traceable SQL, zero data movement
• Barrier removal drives 3.7x ROI and 40% productivity gains
• Healthcare demo setup: hospital COPD readmissions, ~$10K cost per incident
• Four design principles: traceability, no data movement, deterministic-first, governance as code
• Main orchestrator agent plans, writes SQL, calls Teradata MCP server
• Complex questions escalate to a context-isolated data scientist agent
• Built on Claude Agent SDK, running Bedrock Claude Sonnet/Haiku/Opus
• Live demo: COPD readmission rates explored through iterative agent reasoning
• Delegation demo: data scientist agent runs in-database analysis, surfaces factors
• Pre/post tool hooks log every step and cost to CloudWatch
• Agent hosted on Amazon Bedrock AgentCore, fully serverless and scalable
• AgentCore delivers runtime, memory, identity, and observability out of the box
• Lessons learned: guardrails first, deterministic ops, multi-agent registry, ongoing evaluation

Participants:

• Trinath Yarlagadda – Principal Solution Architect – Agentic AI, Teradata
• Neha Wadhera – Sr Solutions Architect, Amazon Web Services  

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

Learn how Vercel's "self-driving infrastructure" vision pairs with AWS databases to eliminate backend friction, securely cutting Aurora Serverless creation time from minutes to seconds.

Topics Include:

• Hedieh Zandi (Vercel) and Manbeen Kohli (AWS) introduce prompt-to-production session
• Vercel powers 18 million developers, maintains Next.js and AI SDK
• Vercel's agentic infrastructure runs on AWS Lambda, CloudFront, and S3
• AI now generates frontend, APIs, and workflows for small teams
• Backend friction remains: credentials, provisioning, database configuration still hard
• Vercel envisions "self-driving infrastructure" that adapts automatically to apps
• New AWS partnership brings native Aurora DSQL and Postgres integration
• Manbeen explains databases now built into Vercel Marketplace and v0
• Aurora Serverless database creation sped up from minutes to seconds
• Aurora Postgres, DynamoDB, and DSQL scale prototypes without rewrites
• Pre-configured templates help builders start RAG or shopping AI apps
• Database security uses OIDC and IAM tokens, no stored passwords
• AWS chosen for agents: low latency, autonomy, one-click simplicity
• skills.sh gives agents reusable instructions, mirrors AWS Kiro's "powers"
• v0 lets users build full-stack apps using natural language prompts
• v0 uses Bedrock models and deploys directly on Vercel infrastructure
• Live demo: v0 builds restaurant app, provisions database, adds Stripe checkout
• Demo ends at AWS console; Rauch quote and hackathon close session

Participants:

• Hedieh Zandi - Product Lead, Vercel
• Manbeen Kohli - Director of Product Management, Aurora and RDS Databases, Amazon Web Services  

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

Kaltura's Ruthie Eisenberg and Yair Neumann reveal how the video giant is reinventing itself as an agentic digital experience company built on AI avatars and hyper-personalized content.

Topics Include:

• Kaltura founded 2006, went public on NASDAQ in 2021.
• Kaltura reinventing itself from video company to agentic digital experience company.
• Shift from static content delivery to hyper-personalized conversational experiences.
• Partners and customers now demand intelligence, not just video infrastructure.
• Kaltura's mission: powering agentic experiences across customer and learner journeys.
• AWS co-sell motion strengthened as Kaltura runs on AWS AI infrastructure.
• Camille used a Kaltura avatar to scale her own presentations.
• Most enterprise websites bury content behind thousands of static links.
• Kaltura builds personalised web pages on the fly, in real time.
• Over 80% of content users see is surfaced for the very first time.
• Acquisitions of eSelf.ai and PassFactory complete Kaltura's agentic content flywheel.
• PassFactory answers: what should this specific person see next?
• eSelf.ai enables multimodal conversational avatars that guide users emotionally.
• 20 years of behavioral data underpins Kaltura's content intelligence advantage.
• GPU scarcity and compute costs shape every AI architecture decision Kaltura makes.
• Kaltura optimises model tiers — strongest for planning, lighter models for execution.
• Fidelity, speed, and cost form a constant triangle in every AI product decision.
• Go-to-market and product teams now work closer together than ever before.
• Pricing shifting from seat-based SaaS to consumption and outcome-based models.
• Kaltura co-creating pricing frameworks with customers across different verticals.
• Internal product agent now handles research, stories, and data analysis autonomously.
• Small two-to-three person squads move fastest in the current AI environment.
• Yair's advice: fail at least once a week, succeed once a quarter.
• Kaltura scaled its CEO via avatar for a live investor earnings call.
• Ruthie's advice: keep the customer at the centre of every single decision.

Participants:

• Ruthie Eisenberg – Vice President, Strategic Partnerships, Kaltura
• Yair Neumann – Senior Vice President of Product, Kaltura
• Kamil Davidov – Sales Leader Israel ISV-BizApps, Amazon Web Services
• Johan Broman – EMEA ISV Head of Solutions Architecture, Amazon Web Services

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

With 250,000 customers and $1.2B in revenue, Monday.com's CPTO explains why they threw out their roadmap and rebuilt everything around agentic AI.

Topics Include:

• Daniel Lereya joined Monday.com when it had just 30 people and five engineers.
• He grew the R&D org from five engineers to roughly 900 over a decade.
• Three years ago Daniel became Monday.com's first ever CPTO.
• Monday.com initially approached AI by adding small features across the product.
• They called this early phase "sprinkling AI dust" — helpful but not transformative.
• A pivotal board meeting made Daniel realise AI hadn't changed Monday's core value.
• Monday.com decided to rethink its mission from first principles around AI.
• The new mission: AI agents that actually execute work, not just manage it.
• AI gives businesses an "infinite workforce" regardless of company size.
• Agents can now do hyper-personalised work at a scale humans simply cannot.
• Monday's platform puts agents at the centre, replacing boards and dashboards.
• Shared context and human-in-the-loop handoffs make their agents uniquely powerful.
• Monday ran an "AI month" — pausing the entire 900-person builder org to transform.
• The month rebuilt team mindset and energy, reminding staff of early startup days.
• Monday also ran an "agentic week" where every department built their own agents.
• Finance built agents to automatically match incoming payments to customer accounts.
• Scaling AI adoption internally remains the biggest challenge across businesses today.
• Monday introduced "effective AI" — balancing capability with cost efficiency.
• They acquired voice AI startup One AI to add specialised model capabilities.
• On pricing, Monday shifted to a hybrid seats-plus-AI-credits consumption model.

Participants:

• Daniel Lereya – Chief Product and Technology Officer, Monday.com
• Kamil Davidov – Sales Leader Israel ISV-BizApps, Amazon Web Services
• Johan Broman – EMEA ISV Head of Solutions Architecture, Amazon Web Services

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

PagerDuty SVP Rukmini Reddy explains why AI is making software operations exponentially more complex — and why the companies that learn and recover fastest will be the ones that win.

Topics Include:

• PagerDuty powers critical digital operations for enterprises and AI-native companies.
• Founded by early AWS employees who experienced always-on system failures firsthand.
• The platform evolved from simple alerting into a full operational intelligence platform.
• Complexity exploded with microservices, cloud-native infrastructure, and multi-cloud environments.
• Reliability must be a core value — not an operational afterthought.
• PagerDuty's culture champions the customer above everything else.
• Employee recognition extends beyond sales to celebrate the whole business.
• AI is accelerating software creation but making operations far more complex.
• AI fails differently — silently, unpredictably, with a much larger blast radius.
• Enterprises should leverage their operational history as a competitive AI asset.
• AI-native companies must build operational resilience early, not bolt it on later.
• The winners won't build fastest — they'll learn and recover fastest.

Participants:

• Rukmini Reddy – Senior Vice President of Engineering, PagerDuty

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

From cracked data foundations to multi-agent AI, Starburst Data's co-founder shares hard-won lessons on getting the right data, not just more of it.

Topics Include:

• Matthew Fuller, co-founder and VP of Product at Starburst Data, joins the show.
• Starburst is built on Trino, a fast SQL engine for federated data queries.
• Their platform lets users query data across lakes, stores, and databases seamlessly.
• Governed "data products" give organizations access to their full data estate in context.
• A strong data foundation is essential before any AI use case can succeed.
• AI doesn't create data problems — it exposes the cracks already there.
• Common mistake: assuming everyone in an org defines "customer" or "revenue" the same way.
• More data isn't always better — getting the right data is what matters.
• Customers include HSBC, Comcast, Zalando, ZoomInfo, and DBS, many running on AWS.
• AWS partnership spans technical support, SLA reliability, and proactive product briefings.
• Advice for product leaders: always anchor new technology back to the customer problem.
• 2026 will be defined by specialized multi-agents working together autonomously.

Participants:

• Matt Fuller – Co-Founder, Vice President of Product, Starburst Data

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/

Find out why the world's largest banks and enterprises trust CockroachDB for mission-critical infrastructure, and what a decade of AWS partnership means for the future of cloud-native data.

Topics Include:

• Cockroach Labs makes CockroachDB, a distributed SQL database built for resilience.
• It delivers cloud-native consistency that legacy relational databases simply cannot match.
• The name "cockroach" reflects survivability — it's designed to never go down.
• Target customers include major banks, trading platforms, retailers, and gaming companies.
• AI is forcing enterprises to accelerate database modernization from the board level down.
• AWS has been a foundational cloud partner for Cockroach Labs for a decade.
• The CockroachDB-AWS integration spans EC2, S3, Bedrock, and Amazon Q-Transform.
• AWS partnership shapes both product roadmap decisions and go-to-market execution.
• New partners should educate themselves first — AWS programs are deep and extensive.
• CockroachDB now supports native vector search for RAG and generative AI applications.
• Agentic AI could mean trillions of digital agents demanding real-time data infrastructure.
• Database modernization and AI adoption will only accelerate dramatically through 2027.

Participants:

• Cassie Zimmerman – Senior Director, Global Strategic Partnerships, Cockroach Labs

See how Amazon Web Services gives you the freedom to migrate, innovate, and scale your software company at https://aws.amazon.com/isv/