Share individual MITechCon 2026 sessions with ready-to-use copy, a public mitechcon.net destination, and a UTM-ready partner link.
Built from the current speaker-site generated agenda data and reviewed session social copy where available.
Apache Iceberg gets powerful when you understand the table lifecycle. The value comes from seeing how snapshots, rollback, compaction, and governance fit into the full table lifecycle. "Apache Iceberg Deep Dive Workshop" turns that scenario into a workshop where Lester Martin walks through hands-on work with snapshots, rollback, compaction, and governance basics.
Defenders make better decisions after seeing the attacker's workflow. Most defenders have never seen an attack from the other side of the keyboard. Ken Smith uses "Certified Red Team Fundamentals" to connect the warning signs to a walkthrough of a realistic red-team lab from recon to privilege escalation, keeping the focus on decisions teams can actually make.
Vague requirements turn into brittle architecture fast. How do you transition a sprawling, complex business problem into a clean, resilient software architecture? That is the entry point for "From Chaos to Context: A Hands-On Journey Through EventStorming and Architecture Katas", where Sarah Dutkiewicz shows EventStorming and architecture katas to move from domain chaos to design clarity in practice.
Prompting is only one layer of useful AI. Once an LLM feature reaches real users, the engineering questions move from prompts to behavior, context, risk, and reliability. "LLMs Under the Hood - Applied Engineering Workshop" lets Barry Stahl connect tokens, embeddings, transformers, RAG, and responsible design to the engineering choices behind reliable LLM applications.
AI-powered vision can improve quality control only when it adapts to real production conditions. Quality systems have to respond to variation on the line, not just perform in a clean lab setup. "AI Influence on Reliable Machine Vision" lets Michael Koper connect vision data, AI, and manufacturing quality controls to the conditions where defect detection has to work.
Public agencies need AI that improves services without adding complexity. Governments are starting to use AI agents to improve how services are delivered—from permits and public works to citizen support. Himanshu Goil brings that problem into "Civic Agents: The Rise Of AI In Modern Public Service" by focusing on practical Civic Agent patterns for pilots, playbooks, and scale, not just the tool names around it.
Autonomous vehicles expand the attack surface across AI, connectivity, and control systems. Cybersecurity threats are not static; it’s a dynamic environment and constantly changing. In "Driving into the Future: Navigating Emerging Cybersecurity Threats in Autonomous Technology", Hemanth Tadepalli moves from the symptom to where emerging risks appear and how to reduce them, helping attendees see where better decisions start.
LLM costs spike when agents send redundant context. With context windows expanding to 200K+ tokens, a single API call can cost several dollars & in production systems handling thousands of requests, these costs compound quickly. The session "Headroom: A Context Optimization Layer for LLM Applications" picks up from there, with Tejas Chopra grounding how context compression and routing can cut token waste without losing useful signal in real implementation choices.
RAG is only as good as the metadata behind it. Large language models are only as useful as the context you give them. "RAG for Data Engineers: What It Is, Where It Fits, and Why Your Metadata Is the Missing Piece" keeps the conversation practical as Varun Joshi works through where schemas, lineage, query history, and dbt models fit in data engineering workflows.
AI agents can damage production data before a human sees the mistake. No confirmation prompt fired, no override path existed, and the database changed before anyone could intervene. In "Your AI Agent Just Ran DELETE Without WHERE. Now What?", Jonathan Stewart turns that failure mode into infrastructure-level gates that block destructive commands at execution time.
Serving an LLM on Kubernetes has sharp edges. Includes a live demo deploying an open-weight LLM on Kubernetes and serving inference traffic with real-time GPU metrics. That is the entry point for "A Practical Guide to Running LLMs on Kubernetes", where Rohit Mishra shows how to handle GPU scheduling, model weights, health checks, autoscaling, and live inference in practice.
Synthetic USB models miss the messiness of real devices. Verifying USB Host designs-under-test (DUTs) is increasingly difficult as the USB ecosystem continues to expand across device types, protocol modes, and class behaviors. "Accelerating USB Host Verification Using Real USB Devices in a Virtualized Device Modeling Framework" turns that scenario into a session where Suchir Gupta shows how physical hardware can accelerate host verification in virtualized environments.
Enterprise prompts break down when too many rules compete for attention. The issue is not that LLMs are weak; reliability changes once instructions, policies, exceptions, and governance collide inside one workflow. In "Engineering Beyond the “Rule of Eight”", Harsh Ranjan turns that failure mode into architecture patterns for governed AI pipelines that hold up beyond a fragile prompt.
LLM features can become new attack surfaces. As AI systems move into enterprise applications, prompt injection, extraction, and jailbreaks turn from demos into real design risks. Ethan DeMott uses "Fortifying AI: Understanding Prompt Attacks and How to Defend Against Them" to connect those attacks to layered defenses teams can design before the feature is exposed.
Digital tools stall when standard work and operators are disconnected. The idea only matters if it survives real equipment, operators, data quality, and production pressure. Bryan Bauw uses "From Standard Work to Scalable Smart Manufacturing: Making Data and AI Work in Assembly Operations" to show how MES, real-time data, and AI can scale assembly improvements without losing the human workflow they are meant to support.
Analytics teams need scale without losing flexibility or control. The session will also cover different ways to ingest data into BigQuery and best practices, and will conclude with a live demo of the BigQuery console and a showcase of a petabyte demo. In "Google BigQuery - Heart of Smart Analytics toolset on Google Cloud", Vijaykumar Jangamashetti makes the problem concrete by showing how BigQuery supports ingestion, ML, and petabyte-scale analysis on Google Cloud.
Sepsis care depends on fast signals across fragmented workflows. The hard part is keeping trust intact as volume, ownership, definitions, and expectations change. In "AI and Digital Transformation in Sepsis Management for Smarter Healthcare Delivery", Swapna Chimanchodkar makes the next step concrete: how AI and predictive analytics can support earlier detection and response.
Conversational BI fails when answers are not trusted. Self-service analytics only works when the data, metrics, and context behind each answer are governed. In "Designing Conversational BI: How Azure Databricks AI/BI Genie Unlocks Self-Service Insights", Mou Rakshit connects Azure Databricks AI/BI Genie to the governance patterns that make self-service insights credible.
Medallion architecture is harder under HIPAA. Bronze, silver, and gold layers are easy to sketch, but regulated healthcare data adds PII handling, lineage, access control, and audit pressure to every step. In "Medallion Architecture in Regulated Healthcare: A Hands-On Playbook", Raziullah Khan turns that architecture into patterns teams can use for cleaner, safer, audit-ready data work.
Most production code is not going to space, but reliability still matters. Ever wonder how NASA writes software that literally can’t afford to fail? The answer gets practical in "Mission-Critical Code: What NASA’s Power of Ten Can Teach Us", where Jonathan "J." Tower shows how NASA's Power of Ten can simplify everyday software design.
AI security changes when systems answer in probabilities instead of certainties. When I say "probabilistic" outcomes, your mind likely jumps to coins or dice. "Probably Secure: What We Misunderstand About AI And Determinism Is Making Us Less Secure" keeps the conversation practical as Dwayne McDaniel works through where deterministic controls still matter and where they fall short.
5G promises low latency, but Android stacks can still bottleneck real-time apps. Real-time mobile experiences need the network, operating system, and edge architecture to work together. "5G-Optimized Android System Architecture for Low-Latency Applications" gives Patel Riddhi a way to make the issue concrete through architecture patterns for faster mobile and edge workloads.
AI agents fail when teams start with the tool instead of the workflow. Once the idea reaches a real workflow, the hard questions shift to context, trust, cost, and reliability. Reshmika Dhandapani uses "AI Agents in Products: A Product Manager’s Guide to Building What Matters" to show how product managers define boundaries, value, and reliability.
Physical pentests should produce more than war stories. Physical penetration testing remains one of the most misunderstood and inconsistently executed disciplines in offensive security. In "Anatomy of a Physical Penetration Test: Recon to Remediation", Ken Smith maps the problem to a repeatable path from recon and pretexting to evidence, reporting, and remediation, so the path from concern to action feels more concrete.
Running many AI models without a gateway creates duplicated auth, weak rate limits, and hidden costs. We cover model routing with A/B traffic splitting, token-based rate limiting per team, fallback chains, and per-team cost attribution. "Building an AI Inference Gateway on Kubernetes: Model Routing, Rate Limiting & Cost Control" turns that scenario into a session where Rohit Mishra walks through gateway patterns for routing and control.
Multi-agent systems get slow and expensive when every agent sees everything. Once the idea reaches a real workflow, the hard questions shift to context, trust, cost, and reliability. In "Multi-Agent Without the Bloat: Designing Lean AI Systems on Azure", Carey Payette ties Azure patterns for scoped context, handoffs, memory, and routing to leaner systems that are easier to debug and scale.
Telecom billing gets messy as services become real-time and partner-driven. This capability allows telecom providers to proactively detect revenue leakage, optimize pricing strategies, and resolve billing issues before they impact customers. Balu Chavan uses "Future of Agentic AI in Telecom Billing and Revenue Management" to connect the warning signs to the underlying mechanics, showing how agentic AI can improve support, revenue accuracy, and leakage detection.
Physical computing gets real when sensors, motion, firmware, and cloud have to work together. When LEGO Meets IoT is a deep dive into building a real-world embedded system using LEGO as the mechanical platform and Arduino as the control layer. The session "LEGO + IOT + Azure = fun!" picks up from there, with Jim Rieck grounding how a LEGO and Arduino build reveals IoT design tradeoffs in real implementation choices.
AI-era security demands more than point tools. In an era where cybercrime rivals the largest global economies, securing digital landscapes has never been more critical. In "Microsoft’s Approach to Comprehensive Security in the Age of AI", David Giard moves from the symptom to the decisions underneath, showing how Microsoft frames identity, data, devices, and threat protection across a modern security platform.
Technical leadership can narrow who gets heard and who advances. But, this is not just a talk about enterprise architecture—it is a conversation about courage, purpose, leadership, and creating space for more people to thrive in technology. That is the entry point for "Three Lessons from a Woman in Enterprise Architecture", where Sarah Wimberley shows three lessons for building curiosity, authenticity, and courage in architecture in practice.
Static routing struggles when latency, cost, and reliability change in real time. Modern distributed systems increasingly rely on real-time decisions across complex service ecosystems. In "Adaptive Machine Learning for Real Time Transaction Routing in Distributed Systems", Jayaseelan Shanmugam connects adaptive ML to routing choices that respond as conditions change.
Finance AI breaks when metrics and pipelines are not trusted. Before finance can safely use AI or agentic workflows, the data foundation has to be trusted, governed, and explainable. The answer gets practical in "Building an AI-Ready Finance Lakehouse with dbt, Governed Metrics, and Databricks", where Mou Rakshit shows how dbt, Databricks, Unity Catalog, and governed metrics create an AI-ready lakehouse.
AI enablement fails when tools arrive without trust, patterns, or coaching. Once the idea reaches a real workflow, the hard questions shift to context, trust, cost, and reliability. In "Floors, Not Ceilings: Lessons from 6 Months of AI Enablement at a $10B Company", Adam Broadbent makes the next step concrete: what a six-month enterprise rollout changed and measured.
Manufacturing teams need clearer instructions, quality control, and process visibility. Operators need guidance in the moment, not after a quality issue has already moved downstream. Yanesh Naidoo uses "Next Evolution of Advanced Manufacturing with ODIN Digital Solutions" to connect ODIN Workstation to real-time operator guidance and the process visibility advanced manufacturing depends on.
Enterprise AI stalls when a chatbot is treated like an application. Millions of dollars were invested in your company's Gen AI chatbot solution. Sriramprabhu Rajendran brings that problem into "Scaling Enterprise Intelligence with Agentic Orchestration — Beyond the Chatbot" by focusing on orchestration patterns that turn narrow agents into useful workflows, not just the tool names around it.
Agentic AI can be technically correct and business-wrong. Agents need business meaning, not just access to more tools and data. In "Semantic Layers: The Missing Foundation for Agentic AI in Enterprise", Sajitha Sritharan shows why semantic layers, knowledge graphs, and governance are the foundation agents need before they can answer business questions reliably.
Production planning breaks when demand, supply, and capacity shift at once. Modern manufacturing environments operate under increasing complexity driven by demand variability, supply uncertainty, and capacity constraints. In "AI Driven Production Planning Using Decision Intelligence in Manufacturing Systems", Madhav Jayesh Kumar Pandya connects decision intelligence to the real-time tradeoffs behind better production planning.
Computer vision gets harder outside controlled labs. Smartphones are becoming powerful sensing platforms, opening new possibilities for healthcare, accessibility, and human-centered AI. "Computer Vision on a Smartphone: Lessons from Building Clinical-Grade Movement Tracking" turns that scenario into a session where Temidire Adesiji walks through lessons from clinical movement tracking on smartphones, where reliability and trust matter.
LLMs are powerful tools, but entity extraction still has classic CS problems. A tool can produce results while the underlying computer science problems are still very much alive. Milena F and Andy Schmidt use "LLMs are tools, not solutions: Entity extraction as a computer science problem" to show how context, ambiguity, and pipelines shape reliable results, giving attendees a practical way to bring the lesson back to their own systems.
EV battery leaks are hard to find without slowing production. The idea only matters if it survives real equipment, operators, data quality, and production pressure. "Robotic Tracer Gas Sniffing for Leak Localization in EV Battery Packs and Energy Storage Systems" gives Davy Leboucher room to explain how robotic tracer gas sniffing improves localization across complex battery assemblies.
AI tools do not fix broken processes, siloed data, or manual approvals. You bought a thousand Copilot licenses. You sent half the company through a prompting workshop. Balazs Horvath uses "The AI is ready, but you are the problem. How to finally be ready for a real AI." to connect the warning signs to what has to change underneath before AI can run real work, keeping the focus on decisions teams can actually make.
Coding agents waste context when they read entire files. Every AI coding agent I've used does the same thing: it reads your entire file, every single time. That is the entry point for "Your AI Reads Entire Files. It Only Needs 6%. Here's How We Fixed It.", where Rajkumar Sakthivel shows how AST-aware retrieval and local vector search can give agents only the code they need in practice.
Some AI decisions need to happen where the data is created. The architecture has to account for devices, models, telemetry, and observation close to where the signal starts. In "AI on the Edge", Jared Rhodes moves from the symptom to a repeatable path for choosing, testing, shipping, and observing edge models, helping attendees see where better decisions start.
Not every software problem needs another neural network. Nature-inspired algorithms can solve certain engineering problems elegantly when teams know where they fit. Barry Stahl brings that perspective into "Amazing Algorithms for Solving Problems in Software", focusing on algorithms that help developers reason through real software problems instead of reaching for the trendiest tool.
Multi-agent AI gets harder when agents need to coordinate at scale. Coordination patterns matter once agents have to share state, communicate, and keep work moving across a system. Mary Grygleski uses "Building Scalable Multi-Agentic AI Systems: Orchestrating Agents with Event-Driven Approach" to follow that thread into event-driven patterns for communication, memory, workflow state, and MCP, giving attendees a practical way to bring the lesson back to their own systems.
Public-sector AI works best when it fits existing workflows. Public agencies have already experimented with basic AI tools, and many are now asking a more meaningful question: how can AI actually support the work that public staff do every day? The answer gets practical in "Civic Agents in Practice: The Next Phase of AI for Government Workflows", where Himanshu Goil shows how Civic Agents can help staff review requests, summarize cases, and reduce repetitive steps.
Slow lakehouse queries often come down to table design and maintenance. While we focus on Apache Iceberg, the techniques apply broadly to Delta Lake and Apache Hive as well. "Optimizing Your Apache Iceberg Lakehouse" keeps the conversation practical as Lester Martin works through Iceberg tactics for stats, compaction, partitioning, and pruning.
Security concepts click faster when you see an attack unfold. A live attack chain makes the gap between theory and defense much easier to spot. In "Real World Cyber Attacks & How to Stop Them", Ken Smith turns each step of the attack into a practical defense conversation, giving attendees a clearer way to recognize and reduce real-world risk.
Pipeline failures should not wait for someone to notice a bad report. Every data engineer knows the 2 AM pipeline failure — the one nobody notices until Friday's report is wrong. "Agentic Loops in the Data Stack: From Pipeline Failure to Auto-Remediation" turns that scenario into a session where Varun Joshi walks through agentic patterns for monitoring, root cause analysis, and auto-remediation.
Static rules create false positives and miss changing crime patterns. The hard part is keeping trust intact as volume, ownership, definitions, and expectations change. In "Applying Machine Learning and Network Analytics for Modern Financial Crime Detection", Shivam Tiwari makes the next step concrete: how ML, behavioral profiling, and network analytics improve detection.
Modernization stalls when teams chase technology before value. When dealing with a 15-year-old platform, the question isn’t just how to modernize but what to modernize. The session "Blending Product Thinking with Software Modernization" picks up from there, with Brian McKeiver grounding how product thinking helps decide what to refactor, rebuild, or leave alone in real implementation choices.
Creative teams do not happen by accident. The best software doesn't come from the best individuals. For teams facing that pressure, "Diverge, Converge, Speak Up: The Discipline of Team Creativity" gives Tom Rodgers room to unpack how psychological safety, decision discipline, and healthier disagreement improve software outcomes. The result is a more concrete path from concern to action.
Your architecture mirrors how your teams communicate. To ignore Conway’s law is a trap that will lead to software that is more expensive, and less effective, especially in the micro-services and composable era. Luis Fernandez uses "Don’t Let Your Org Chart Write Bad Code" to connect the warning signs to the underlying mechanics, showing how Conway's Law and Team Topologies can reduce silos and improve software design.
Municipal infrastructure needs security that respects always-on operations. The risk is easiest to reduce before it becomes an incident, so the workflow matters as much as the warning sign. In "Securing Cyber Physical Systems in Municipal Infrastructure Environments", Venkata Kartheek Reddy Somasani makes the next step concrete: segmentation and field-tested controls for OT/IT cyber physical systems.
Enterprise AI gets simpler when data and models work closer together. Oracle AI Database 26ai introduces a unified, enterprise‑grade foundation for building AI‑powered applications directly where your data lives. In "Building Enterprise-Grade AI Solutions with Oracle Generative AI Service and the Oracle AI Database", Juarez Junior moves from the symptom to the decisions underneath, showing how Oracle AI Database and OCI Generative AI support secure, scalable apps.
Enterprise investigations slow down when evidence is scattered across systems. The challenge is bringing retrieval, agent workflow, and human oversight together without burying investigators in more noise. In "Case Study: Designing AI-Augmented Enterprise Investigation Systems", Santosh Vasudevan and John Selvaraj Arulappan connect AI agents and RAG to investigation workflows where speed still needs human judgment.
Digital transformation fails when leaders start with tools instead of problems. Most organizations do not fall behind because the technology is missing; they fall behind when decisions, change, and readiness stay fuzzy. "Digital Transformation or Digital Chaos? A Practical Guide for Leaders Who Are Done Playing Catch-Up" gives Matt Russell a way to make the issue concrete through a framework leaders can use before another tool rollout turns into noise.
Plant data is valuable, but industrial devices and old protocols make collection intimidating. The first useful step is often less about dashboards and more about getting trustworthy signals out of the equipment already on the floor. Grant Pinkos uses "How to get started with collection and storage of industrial data" to make that starting point concrete, from collection and storage to analysis and alerts.
AI can speed up concepting without replacing creative judgment. Once the idea reaches a real workflow, the hard questions shift to context, trust, cost, and reliability. The answer gets practical in "Knights, Castles, and Creative Cloud: Rapid Illustration Concepting with AI", where Rex Rainey shows how Adobe Firefly supports faster iteration while leaving room for craft.
Manual testing slows delivery and weakens release confidence. The testing strategy has to scale across browsers and platforms without burying teams in maintenance work. "Test Smarter with AI: Playwright Automation at Scale on Azure" keeps the conversation practical as Randy Pagels connects AI-generated Playwright tests with Azure App Testing for broader, faster coverage.
Data pipelines rarely fail on day one. The failure usually arrives later, when schema drift, volume growth, and ownership changes expose early shortcuts. Chris Birie uses "Designing Data Pipelines That Don’t Hate You Six Months Later" to follow that thread into design patterns for schema change, idempotency, observability, and growth before handoff gets painful, giving attendees a practical way to bring the lesson back to their own systems.
Modern products rely on operations customers never see. Delivery, reliability, data trust, and cloud spend all depend on disciplines that become visible only when one gap slows the business. In "The Invisible Stack: Black-Ops of Modern Enterprise", Robert Gray III maps DevOps, AIOps, DataOps, and FinOps as one operating system for modern enterprise work, showing why the invisible stack matters before delivery, incident, or cost problems surface.
