DevWorld 2026
who it's for
software engineers, platform engineers, engineering managers, machine learning engineers, site reliability engineers, security engineers, developer experience engineers
topics
agentic AI architecture; AST-driven code fixes; model context protocol; retrieval augmented generation with knowledge graph; AI SDLC and governance; supply chain slopsquatting attacks; GitOps workflows with FluxCD; serverless functions for sovereign cloud; React loading performance; production hook instrumentation for React; minimal container images with Wolfi OS
| title | author | length |
|---|---|---|
| What the F is Going On? Cognition is Cheap, Understanding and Judgement is Not. | Harijs Deksnis | 23m |
Argues that cognition becomes abundant so generation of facts and code is cheap, but understanding and judgment are scarce; uses isochronic maps and a y = 3x +… | ||
| Stop Chat-and-Hope and Start Building Better Apps with Agents | Salih Guler | 32m |
Live demo builds a production‑grade fullstack app that monitors Hacker News front page, reads top‑comment threads for the top five stories, and emits a… | ||
| AST + AI = Developer Co-Pilot 2.0 | Andrei Tazetdinov | 26m |
Combining ASTs with LLMs produces concrete, actionable fixes instead of generic advice: code is parsed and navigated to exact functions, transformations are… | ||
| AI-Native: The Next Revolution after Cloud Native | Pini Reznik | 22m |
Claim: AI cuts development cost toward zero and lets non‑developers deliver useful apps — a CFO built an internal app with zero dev background and a site was… | ||
| Agentic AI Meets Event-Driven Architecture | Harj Rangarajan | 29m |
Claims that event-driven architecture turns agentic AI adoption from a brittle integration problem into a manageable platform: 20% of the work is models, 80%… | ||
| Trust issues: Because Zero Trust Isn't Optional Anymore | Jan Peer Stocklamir | 25m |
Claims that React2Shell lets an attacker send a malicious React server component, execute it and get a credentialless reverse shell because outbound traffic is… | ||
| The Industry Standard hoax: How Engineers Can Steer Product Decisions | Petar Grchevski | 17m |
Claims that the industry standard hoax matters: engineers get concrete tactics to challenge "we must copy Amazon" thinking, compare multiple real examples, and… | ||
| How to Structure Context to Make Your Agents Smarter | Nyah Macklin | 29m |
Claim: agents fail from fractured context not bad models, so add a structured knowledge layer to retrieval. Demonstrated architecture pairs vector RAG with a… | ||
| How can we abuse AI hallucinations to feed the already bloated software supply chain | Nigel Douglas | 26m |
Claim: generative models hallucinate plausible package names and attackers register those slopsquatting packages on PyPI DockerHub and Hugging Face to push… | ||
| Startup Pitch - Yesglot | Efe Öge | 12m |
Claims an AI-first agentic translation platform named Yesglot that creates PRs to your repo by installing a GitHub app and adding a Yesglot.yml/toml config. It… | ||
| You say you've got an "AI SDLC" | Alexey Krivitsky | 34m |
Claims that an AI SDLC exists get narrowed to specifics: Fabian Wesner runs a 50‑page spec against new models, testing end‑to‑end e‑commerce delivery and… | ||
| The AI-First Playbook: Building and Running a Business with AI | Vincent Jong | 28m |
Concrete specifics: built four companies in 12 months — MeetKite (Calendly alternative), DataMerch (contact-data product), Scope (nontechnical OpenClaw… | ||
| AI Org Change Management: Team Transition to the Agentic SDLC and Ways of Working | Volodymyr Malyk | 38m |
Case study of a Booking.com rollout claims a working spec‑driven multi‑agent pipeline that turns meeting transcripts and docs into LLM-Ready digital footprint… | ||
| Agents at Work: How GitLab Is Redefining AI Adoption for Engineering Teams | Viktor Van Den Berg | 31m |
Claims that agents execute event-driven workflows inside GitLab and actually ship code: the Duo Agent Platform reads an issue, runs an issue-to-MR flow… | ||
| Measuring AI Adoption Across People & Orgs (Maturity Model) | Roman Rodomansky | 17m |
Builds and ships an open-sourced maturity model that uses Google Forms surveys, a six‑part scoring system, simple dashboards and a small stack (Superbase +… | ||
| Giving AI Agents a Wallet | Kevin Lewis | 25m |
A working HTTP payment loop appears: HTTP 402 returns structured metadata, the client signs an EIP‑3099 authorization, a facilitator submits an on‑chain USDC… | ||
| The Model Context Protocol: A Deep Dive into the New AI Security Attack Surface | Ricardo Goncalves | 27m |
MCP standardizes model‑to‑tool integration and claims rapid uptake—uploader cites Antropic 97 million+ downloads per month across SDKs with 10,000+ public… | ||
| Beyond Hiring: Why Women Leave Tech—and How to Build Teams Where They Stay and Lead | Magdalena Szumna, Gabrielle Tang, Nikhita A. | 22m |
Claims a startling metric — 50% leave by mid‑career — and links causes to repeated microaggressions, biased promotion and visibility gaps rather than… | ||
| Serverless by Design: The Role of Functions in Europe’s Sovereign Cloud Future | Christian Boelle | 23m |
European sovereign cloud offers a FunctionGraph service on T‑Cloud (Deutsche Telekom T‑Systems) that mirrors Lambda with API Gateway, OBS (S3), DMS (Kafka)… | ||
| Building Secure Services with AI Agents: Practical Patterns | Marat Kenzhebulatov | 19m |
Greenfield payments backend gets built whole service in two days using AI agents but agents introduce regressions and violated PCI DSS checks during code… | ||
| Unlocking AI Agentic UX with Google’s AP2 Standard and the x402 Protocol | Ben Greenberg | 24m |
Claims that agents need native payment rails and shows a working stack combining x402 payment responses with AP2 authorization and verifiable digital… | ||
| Culture Is Strategy: Leading Tech Teams Across Cultures | Kemo Camara, Magdalena Szumna, Gwen Kolader | 28m |
Claims: culture becomes a measurable growth lever via cultural intelligence; tools: create a common language, build recruitment and retention processes, use… | ||
| AI Agents in Production: Why the Platform Matters | Luuk Hofman | 24m |
Claims that agents need persistent memory, durable execution, real‑time comms and security, supported by concrete primitives: Workers on V8 isolates with… | ||
| Scaling for Intelligence: Distributed SQL Databases in the Age of AI | Wei Hu | 27m |
Claims concrete outcomes and tools: uses Oracle Globally Distributed Database and distributed SQL patterns like sharding, raft quorum replication, and vector… | ||
| Push to Prod: GitOps Workflows on k8s with FluxCD | Niels Groeneveld | 22m |
Claims a pull based GitOps flow where Flux bootstrap installs six controllers and a source controller + Helm controller pair does continuous sync and lifecycle… | ||
| AI-Native Databases: Four Innovations That Matter | Wei Hu | 27m |
Argues that a single Oracle AI Database replaces polyglot sprawl by hosting relational, JSON, graph and vectors in one engine, claims transactional consistency… | ||
| It's Time to Talk About Signals | Julian Burr | 30m |
Explains that signals are reactive variables, shows a working simple JavaScript demo (button counters, computed sum and parity) and implements a minimal… | ||
| Engineering Leadership Through a Social Network Lens | Gabriel Ramirez | 25m |
Scripts on GitHub pull GitHub interactions into a CSV edge list and run every two weeks to compute nodes, density and mean node connectivity; Neo4j plus LLMs… | ||
| Decoding Javascript: Loading Performance | Arvin Abedinlou | 23m |
Starts from a 3.05 MB first-chunk (not gzipped) problem and shows concrete fixes: analyzing bundles with webpack/Rollup artifacts, fixing a module side effects… | ||
| Beyond Websites: Creating Web Spaces | Lara Brook | 16m |
Claim: the web already supports building interactive web spaces not just pages using existing browser APIs. Specifics: phones expose nine sensors, web speech… | ||
| The Inbox is an Agent Runtime | Confidence Okoghenun | 30m |
Argues that the inbox is an agent runtime and that email is the best protocol because it supplies identity, durable threaded context, and async wake events… | ||
| Automating Accessibility with Playwright | Rowdy Rabouw | 28m |
Demonstrates automating accessibility scans using Playwright with Axe Core and the axe‑html‑reporter, sets up via npx playwright (choose TypeScript, install… | ||
| Models Gone Rogue | Elena Calvo | 23m |
Covers four concrete failure modes and fixes: prompt injection with examples of direct and indirect attacks and mitigations like Zod validation and the… | ||
| The Formula for Faster Outage Recovery | Maxim Scheplin | 24m |
Claims a simple formula: outage = time to detect + time to acknowledge + time to repair and shows concrete levers for each. Defines SLO examples like 99% of… | ||
| The New Frontend Stack: Humans, AI, and Prompts | Andrei Tazetdinov | 17m |
Creates a React TypeScript invoice form and runs strict unit tests against several LLMs (Mistral, Code Llama, Gemini, Claude Sonnet), reporting initial pass… | ||
| Rust-Inspired Typescript | Rijk van Zanten | 30m |
Rust habits mapped to concrete TypeScript patterns: make impossible states explicit with discriminated unions and exhaustive never checks, return Result-like… | ||
| How Great Tech Teams Build a Productivity Culture that Scales | Arinc Kokturk | 31m |
Claims a method to keep productivity high while scaling by designing workflows, rotating roles, and treating AI as a team member. Uses Claude to generate tests… | ||
| Building and Deploying A Multi-Agent AI D&D with Typescript | Salih Guler | 28m |
Builds a working multi‑agent Dungeons & Dragons demo that actually creates characters, rolls dice, queries rules from a PDF, and runs a front end—code is… | ||
| Under the Hood of Undici | Carlos Fuentes | 24m |
Undici is a from‑scratch HTTP client that swaps Node’s event‑emitter model for a dispatcher state machine, uses a Wasm llhttp bridge instead of native… | ||
| Beyond Bugs: How AI turns QA into Full-Circle Engineers | Alex Shershebnev | 22m |
Claims that AI makes QA full engineers get grounded in specifics: Zencoder’s ZenFlow runs an agent fleet, a PR test-coverage bot suggests P1/P2 scenarios and… | ||
| 10 Hadar | Hadar | 19m |
A team builds an internal retrospective app in React TypeScript Tailwind over six months — a few screens and a few thousand lines of code — by vibe coding in… | ||
| The time paradox: Building Timezone-safe python/django applications. | Denny Biasiolli | 27m |
Root cause: UTC-stored timestamps plus per-user local‑time expectations and staggered DST rules cause missing Q1 revenue and mis-scheduled jobs. Tools and… | ||
| Sleep better on release day | Ondřej Chrastina | 28m |
Analysis of a 40,000-test repo surfaces practical tactics: run broad coverage across frameworks with deep validation on Vit and Vanilla JS, use Cypress and… | ||
| React Native in Practice | Milica Aleksic | 25m |
Production React Native work reduces to concrete surprises: apps used by a million users break from small repeated choices; a reused card that grows into 23… | ||
| Killing Wasted Re-renders With Production Hook Instrumentation | Andrei Tazetdinov | 13m |
Claims that about half of renders are wasted and demonstrates a concrete fix: build-time Babel AST injection adds production hook insturmentation that wraps… | ||
| I Followed the Best Practices and Still Got Hacked | Irina Iarlykanova | 22m |
Shows concrete failures: a middleware that validates a value from the query string while handlers use the JSON body leading to negative-tip via dual sources… | ||
| UI in The Age of AI | Adam Cowley | 21m |
Neo4j Graph Academy experiments with a 2023 RAG chatbot that goes into production, proves coherent but unhelpful for platform faults, and is quietly removed; a… | ||
| TMCP: A New Way to Build MCP servers in TS | Paolo Ricciuti | 29m |
Claims a minimal TypeScript MCP server library called TMCP that replaces the official SDK with a lightweight JSON-RPC core using JSON-RPC 2.0, adapters for… | ||
| Hack Me if You Can | Dani Coll | 29m |
Shows a concrete React to Shell 10/10 remote RCE: crafted react flight JSON with __proto__.constructor.constructor executes payload and runs execSync… | ||
| building reliable AI systems in production | Anand Chowdhary | 26m |
Claim: concrete systems for reliability not vague best practices. Shows a dashboard that logs every generation, prompt versioning with sandbox side-by-side… | ||
| My Context Progress - A Psychologists Journey to Implementing an MCP | Bernhard Mayr | 29m |
Storyblok builds an MCP server that maps ~140–160 management endpoints into a curated toolset using open API overlays so the source spec stays untouched, and… | ||
| Simplicity in complexity | Burcu K Kutluay | 26m |
Interactive simulation rebuilds a failing feature found six months after release and walks through concrete tactics: reproduce-debug-pair, split into LEGO-like… | ||
| Creative engineering Becoming the Leader You Want to Be | Pragya Arora | 27m |
Decision process is concrete: use a decision tree — don't know vs don't want — and either upskill or motivate. Coach a quiet engineer who says "Who am I to… | ||
| How to prevent a rewrite | Michiel van der Ros | 24m |
Seven-year monorepo with 35,000 commits, four active developers and deliberately zero rewrites; concrete moves include migrating to Storyblok, adopting… | ||
| Culture at Speed Scaling Without Losing People in the AI era | Shokoofeh Ketachi, Ania Hajdrowska, Kimberly Fuqua | 26m |
Panel claims that everyday work gets offloaded to tools like Copilot ChatGPT Cloudera and that leaders who don't ask questions lose ground; they cite 84% of AI… | ||
| Managing Small Datasets Without Losing Your Mind | Oleksii Liashuk | 23m |
Starts with 2,000 images and ends up managing ~200,000, reduces dataset update time from 5–7 days to about 2 days by using a clear stage pipeline instead of… | ||
| All Your AI are Belong to Us | Alex Shershebnev | 28m |
Claims: generative models introduce web app bugs and miss Supabase role rules leading to an exploit, agents can autonomously pen‑test with tools like Shannon… | ||
| Sustainable IT: The path to a greener future | Saravanan K Nagarajan | 24m |
Smartphones already do trillions of operations per second vs Apollo 11’s ~14,000 FLOPS, and the sector drives 2–3% of global emissions—roughly the same as… | ||
| Risk, Evidence and Conformity | Daniel Thompson-Yvetot | 30m |
The Cyber Resilience Act requires a CE mark backed by traced evidence: keep test artifacts for 10 years, rerun a new declaration whenever a change alters the… | ||
| Self hosted GIS with open geo data - Real-world experience | Mikhail Chechulin | 16m |
A company builds a self‑hosted GIS on OpenStreetMap using ORS + Pelias on Kubernetes and ships a facade API for geocoding and driving‑time queries, completing… | ||
| How Auth0 Secures AI Authentication & Authorization for AI Agents | Jeremy Durand | 29m |
Claims concrete fixes: use a scoped token vault to keep refresh tokens off agents, implement CIBA backchannel for human‑in‑the‑loop approvals so an unattended… | ||
| Worms, Tokens, and Trust The Industrialization of Supply Chain Attacks | Mackenzie Jackson | 29m |
Claims that LLMs can spot fixes and malware faster than CVEs and a working pipeline that scrapes 5 million changelogs and analyzes code drive the talk: 2,106… | ||
| Bots & Humans: Maintaining AI agent and automation in an AI-bot web | Anil Kumar Krishnashetty | 24m |
Claims the web is increasingly closed, citing 199 million requests to docs from Mentify and Cloudflare blocking 400 billion bots, then shows a build: uses… | ||
| Elevating Quality: How AI Enhances Testing in Continuous Delivery Pipelines | Berry Langerak | 28m |
Tilaa describes building an AI pipeline that finds end-to-end coverage gaps by parsing Vue code back from Apollo calls, emitting YAML coverage format, and… | ||
| Building minimal images with Wolfi OS, declaratively with melange and apko | Django Beek | 30m |
Builds a minimal Python image from source with Wolfi OS using melange to produce APKs and apko to compose an OCI image, emits an aggregated SBOM embedded in… | ||
| The parts of DNS nobody explained to you | Nico Kempe | 18m |
Crash course that turns vague DNS lore into actionable facts: delegation from the root zone to ~1,590 TLDs with ~1,400 usable and ~1,200 gTLDs plus ~200… | ||
| Management is hard job! That's why you should do it like an engineer | William Mendes | 29m |
Management becomes an engineering problem: design team architecture, define human API contracts and add observability with SLOs and dashboards. Clarity matters… | ||
| API Productivity by Design | Szymon Drosdzol | 27m |
Claims that API architecture antipatterns create security debt and slow sprints, backed by a real bug where a crafted parameter returns session tokens leak via… | ||
| Rise of Risk Ops teams - A solution for reducing cognitive load on DevOps teams. | Sutapa Sarkar | 27m |
At ING a dedicated RiskOps enabling team takes ownership of monitoring, evidencing and tooling so DevOps stops juggling audits and fixes. Implementation… | ||
| The Backstage Builder Experience at Adyen | Sander Boumeester | 17m |
Adyen runs backstage as a single-portal platform and reports a two‑year build-from-zero journey: React front end, Node.js back end, Postgres, a populated… | ||
| How AI is Re-shaping Engineering Work (And How to Deal With It) | Juliette van der Laarse | 33m |
Presents an AI flower framework that maps 150 IT activities to concrete “North Star” standards, built in Neo4j and linked to Claude to auto-generate… | ||
| Agent Experience | Matt Billman | 33m |
Claims that most web code will be written by autonomous agents and that products must be rebuilt for Agent Experience. Demonstrates concrete tooling: open… | ||
| Fireside Chat on Documentation | Sara Tandowsky | 31m |
Claims that Gitbook sees 50%+ machine access to docs, that AI-assisted writing adoption more than tripled year over year, and that docs are shifting into a… | ||
| AI for Everything After Code | Martin Reynolds, John Crespo | 28m |
Claims concrete outcomes: AI embeds security scans (Semgrep, GitLeaks), surfaces reachable vulnerabilities and AI creates PR fixes for developer review… | ||
| Teaching Agents to pay: A Developer's Guide | Allison Farris | 25m |
Claim: agent-first checkouts already convert better and one in four people (2025) use chat chatbots to discover products. Demo uses Stripe Payment Element to… | ||
| Engineering Productivity Starts with Predictable Behaviour | Srini Srinivasan | 27m |
Argues that engineering productivity follows from measurable predictability: replacing cache‑heavy stacks with an SSD‑optimized NoSQL yields steady low latency… | ||
| Code vs. Compliance | Daniel Friedlaender, Masha Moisseyeva, Mark wolkenfelt | 46m |
Claims that regulation both enables and strangles growth get concrete numbers and fixes: a EU-wide survey finds 80% of founders avoid specific European markets… | ||
| How Web Data Infrastructure Powers the Next Generation of AI | Particija Zemaityte | 21m |
Oxylabs describes building a production web-data layer: a video API in under two weeks handling multi-petabyte jobs, iterating from transcripts to subtitles… | ||
| Fireside chat on Future Developer Experience | Matt Biilmann | 53m |
Claims rapid shift from developer-centric tooling to an agent experience where coding agents do reading and writing of code and humans orchestrate. Specifics… | ||
| Multiplayer Game Dynamics | Shafik Quorasihee | 27m |
CrossPlay is a standalone New York Times' multiplayer app with a 15 by 15 board, seven tiles drawn from a 100‑tile bag, and three bot difficulties; on‑device… | ||
| It passed auth, then production caught fire. | Alex Olivier | 20m |
A concrete failure mode: distributed services implement authorization differently so six services return divergent decisions for the same user and action, one… | ||
| An AI Operating System | Lee Briggs | 22m |
Claim: treat AI as an operating system by centralizing usage and measuring real impact. Built Aperture to proxy models, enforce identity-based access, log… | ||
| nothing matches | ||