The gap between "it worked in the demo" and "it holds up in production" is where most AI work lives. The InfoQ Certified AI Engineering Program covers that gap. First cohort starts July 25. 🔗 bit.ly/3Qybpuq

The InfoQ Certified Architect Program is for senior engineers and architects carrying a decision they’ve been sitting with for a while, finally getting feedback from people with no stake in how the answer lands. Starts August 13. 🔗 bit.ly/4wd9WZu

Past QCon London attendee: "I was learning from what is actually happening in other companies." QCon London 2027. April 13-15. 75+ practitioners, 15 tracks. Current pricing ends July 14th. 🔗 bit.ly/4eF7e9r

Most AI experiments cover the happy path. The edge cases, inputs your eval suite missed, and scenarios that only surface with real users come later, in production. Hien Luu, author of MLOps with Ray, facilitates the first InfoQ Certified AI Engineering Program cohort→ bit.ly/4v6zvL2

𝟐𝟎 𝐲𝐞𝐚𝐫𝐬 𝐨𝐟 𝐈𝐧𝐟𝐨𝐐! 🎂 We're hearing from our editors. 𝐀𝐧𝐭𝐡𝐨𝐧𝐲 𝐀𝐥𝐟𝐨𝐫𝐝, AI, ML & Data Engineering, shares how InfoQ helps him keep up with the state of the art in a field moving this fast. Thanks, Anthony! 🙌 🎂 𝐇𝐚𝐩𝐩𝐲 𝟐𝟎𝐭𝐡 𝐀𝐧𝐧𝐢𝐯𝐞𝐫𝐬𝐚𝐫𝐲, 𝐈𝐧𝐟𝐨𝐐! #InfoQ20 #AI #DevCommunity

As you get more senior, where does outside input on architecture decisions come from? InfoQ Certified Architect Program: 4 hrs/week, 5 weeks, senior peers from other companies. Next cohort August 13. 👉 bit.ly/4ox6QNf

QCon London 2027 is April 13-16. On April 16, Luca Mezzalira runs a half-day architecture certification training. Combined ticket available. Current pricing ends July 14th. bit.ly/4efK8pz

QCon London 2027 registration is open. April 13-16. 75+ senior practitioners, 15 tracks, no hidden product pitches. The speakers have already shipped what you're planning. Current pricing ends July 14th. 🔗 bit.ly/3PTqRkx

𝟐𝟎 𝐲𝐞𝐚𝐫𝐬 𝐨𝐟 𝐈𝐧𝐟𝐨𝐐. 🎂 We're hearing from our editors. 𝐎𝐥𝐢𝐦𝐩𝐢𝐮 𝐏𝐨𝐩, one of our InfoQ editors: a community where senior & junior developers connect, collaborate, and learn from one another. Thanks, @olimpiupop.bsky.social ! 𝐇𝐚𝐩𝐩𝐲 𝟐𝟎𝐭𝐡 𝐚𝐧𝐧𝐢𝐯𝐞𝐫𝐬𝐚𝐫𝐲, 𝐈𝐧𝐟𝐨𝐐! 𝐋𝐚 𝐦𝐮𝐥𝐭̦𝐢 𝐚𝐧𝐢! #InfoQ20

Video and transcript from my talk at QCon London are now online. Thanks @qconferences.com! www.infoq.com/presentation...

𝟐𝟎 𝐲𝐞𝐚𝐫𝐬 𝐨𝐟 𝐈𝐧𝐟𝐨𝐐! 🎂 We're celebrating by hearing from the editors who build our community. Almir Vuk, InfoQ Lead .NET Editor, went from reader to contributor to editor - growing as an engineer and speaker along the way. Thanks, @almirvuk.bsky.social! 𝐒𝐫𝐞𝐭𝐚𝐧 𝐫𝐨đ𝐞𝐧𝐝𝐚𝐧! #InfoQ20 #DevCommunity

Twenty years ago today #InfoQ version 0.7 was officially launched 🎂

Production agents fail underneath the loop: durable state, daily deploys, sandboxes, model portability, observability. The Architecting for Agents track at #QConSF 2026 works through each layer with practitioners shipping agents in production today: bit.ly/43bRm7D

The more senior you get, the fewer people question your architecture calls. The InfoQ Certified Architect Program is the peer group that does. June 10: bit.ly/3PDnnT7

Day 2 #QConAI Boston 🚀 💡 How engineering teams are measuring & growing AI maturity 💡 Multi-agent platforms built for real org complexity (Spotify, OpenAI) 💡 AI infrastructure: GPU markets, LLM ranking systems &cost transparency 💡 What 25 trillion tokens tell us about how developers actually adopt AI

Halfway through Day 1 at #QConAI Boston, a few things are standing out: 🔹 Context is becoming an architecture problem, not just a prompt problem 🔹 Trust and safety at scale needs real infrastructure, not just policies 🔹 KV cache is one of those levers teams notice when inference costs start to hurt

A theme showing up across #QConAI Boston today: production AI is mostly a systems problem now. The model still matters, but the harder questions are around cost, latency, context, compliance, security & scaling AI work beyond a few specialists: boston.qcon.ai/schedule/bos... #AIConference

🎤 Kicking off Day 1 #QConAI Boston with keynote from Martin Spier, Engineering Leader OpenAI: "Keeping ChatGPT Fast in the Agentic Era." Martin dives into the infrastructure and systems thinking behind keeping one of the world's most-used AI products fast at scale: shorturl.at/MctwH #AIConference

Day 1 of #QConAI Boston 2026 is ON! 🎉 Two days of practical, production-focused AI, real systems, real teams, real constraints. Let's explore what practitioners are doing to ship AI safely, scale inference reliably, and build the agentic systems that actually hold up in production: boston.qcon.ai

Big technical transitions get carried by Staff+ engineers. What that work looks like in practice: the Staff+ Engineering track at QCon San Francisco 2026. bit.ly/3Q7EpJ2

Someone has to make the platform call, ship the first agent, and build the eval strategy that holds under real traffic. The InfoQ Certified AI Engineering Program is for staff engineers, engineering managers, and ML platform leads. Starts July 25, 2026. → bit.ly/4tT2HEp

Platform and value-stream design are hard to check from inside your org. The Organizational Architect cohort with Manuel Pais opens June 19: bit.ly/3PoYwlS

Your AI prototype worked. Then real users showed up with questions, load, and edge cases you didn't anticipate. The InfoQ Certified AI Engineering Program starts July 25, 2026. 5 weeks online with Hien Luu, author of MLOps with Ray. 🔗 bit.ly/43q0yoT

When ChatGPT feels slow, users blame the model. The real bottleneck is usually something else entirely. Martin Spier (Head of ChatGPT Performance @OpenAI) on full-stack AI latency and agents that investigate their own regressions. QCon AI Boston, June 1–2: bit.ly/3Rce3WH

~90% of enterprise AI agents stall at POC. The problem: no systematic way to know when an agent is ready. Zhou (Jo) Yu (CS Professor @Columbia, Founder @Arklex.ai) on multi-turn simulation and eval methods that break the demo loop. QCon AI Boston, June 1–2: bit.ly/4dWzxi3

Training days are available at QCon San Francisco 2026 (November 19–20), right after the conference. Hands-on workshops from senior practitioners. Smaller groups, focused topics, time to actually work through the material. Add to your pass or attend standalone. Spots are limited: bit.ly/4uSEFup

Agents fail in production not because of the model but because of what's missing around it. Vinoth Govindarajan (MTS @OpenAI) on control planes, single-writer execution, and approval boundaries that keep agents predictable under load. QCon AI Boston, June 1–2. Find out more: bit.ly/4dqq4Pz

Spotify's ad platform: audience targeting, creative gen, campaign goals, and budget optimization. A single agent for all of it = slow and hard to debug. Pratik Rasam, Senior Engineer @Spotify, on rebuilding it as a multi-agent pipeline. QCon AI Boston, Jun 1-2: bit.ly/4dFpFdi

They built an agent system with a prompt-chain framework. It demoed well. Then real workloads hit: tool calls lasting hours, workflows that needed to survive restarts. They threw it out and rebuilt on the actor model. Manju Rajashekhar on the patterns that survived: bit.ly/4wf7P8l

Fabiane Nardon, Data Platform Director @TOTVS, on the data layer beneath reliable agent systems, and the token economics that quietly kill most projects. QCon AI Boston, June 1-2. Current pricing ends May 12. 🔗 bit.ly/4cPGlP2

The QCon San Francisco hotel block at the Hyatt Regency is already 65% booked, six months out. Staying at the venue keeps you in the conversation between sessions. Nov 16-18. Early bird May 12. 🔗 bit.ly/4dn2iFk

When an agent approves a loan or classifies risk, the output can't be probabilistic. Alex Porcelli on how a financial services enterprise runs decision models as agent skills, processing millions of evaluations daily. QCon AI Boston, June 1-2. 🔗 bit.ly/3P3UsqH

Four hours a week sounds like a lot. Architects who've done the InfoQ Certified Architect cohort say it replaced unstructured thinking they were already doing, with external feedback they couldn't get internally. May cohort starts May 7. Registration closes May 6. 🔗 bit.ly/3PlfzF9

Rachel Shalom, Distinguished Engineer @Dell, on embedding models, diagnosing failure modes and when fine-tuning is worth it, from Dell's agentic search system. QCon AI Boston, June 1-2. Current pricing ends May 12. 🔗 bit.ly/3PaPBUM

QCon San Francisco 2026 tracks are out. 12 tracks, 60+ speakers. Architecting for Agents, Guardrails and Safety Nets (evals), Staff+ Engineering Skills, Developer Experience as an architecture concern, and more. Nov 16-18. Current pricing ends May 12. 🔗 bit.ly/4t8DntM

Forter gave 200 people two weeks to build their own AI agents. They built 80. Engineers, PMs, analysts. Ben Maraney, Principal Engineer @Forter, on the minimum viable platform and why their MCP server was the biggest force multiplier. QCon AI Boston, June 1-2. 🔗 bit.ly/4d7zEb5

"I go to companies that are over-engineering their systems because they're not asking the right questions." Luca Mezzalira on what the in-person Certified Architect cohort at QCon San Francisco covers: ADRs, team topologies, trade-off mapping, and Architecture Decision Forums. 🔗 bit.ly/4tTNR0M

Red Hat AI's 500-engineer org is adopting agents across their SDLC. They started measuring speed. Their engineers pushed back. Catherine Weeks on what they measure now: including a security pipeline that took 4 iterations to go from 58% false positives to 22%. 🔗 bit.ly/4cs8bkk

Partial schedule for QCon AI Boston (June 1-2) is live. Sessions from DoorDash, LinkedIn, Apple, Red Hat on agents, context engineering, AI safety at marketplace scale, and more. Early bird ends May 12. 🔗 bit.ly/4euJi8L

"Code is becoming a commodity. Good code for your context isn't." Luca Mezzalira on why every senior engineer now needs to think like an architect. A full week of the cohort covers it. May and June cohorts open: bit.ly/4u3YT3T

AI agents don't know your company. Ajay Prakash, Senior Staff @LinkedIn, on how CAPT gave 500+ skills to engineers, PMs, and designers via MCP, with a live demo. QCon AI Boston, June 1-2. Early bird ends May 12. 🔗 bit.ly/48He2jc

Direct LLM-to-SQL doesn’t hold up at enterprise scale. Francesca Lazzeri (Microsoft) shares a hybrid architecture that combines LLM reasoning with deterministic computation for reliable query generation. QCon AI Boston, June 1-2. Register: bit.ly/4dQ8iHl

Reliable LLM outputs in production require more than prompting. Ricardo Ferreira (Redis, ex-AWS, Elastic, Confluent) covers context engineering as an architectural discipline at QCon AI Boston, June 1-2. Register: bit.ly/484fBYf

Early in your career? The QCon AI Boston volunteer program offers free access to sessions and recordings in exchange for onsite support. June 1-2, Boston. 🔗 bit.ly/4tRwvC1

Moving an AI agent from prototype to production introduces state management and framework lock-in problems most tooling wasn’t built for. Deepak Chandramouli and Bhumik Thakkar @Apple share how they solved it with Ray. QCon AI Boston, June 1-2. Early bird ends today. 🔗 bit.ly/4elqp88

Registration for the April architecture online cohort closes Monday, April 13. Join Luca Mezzalira and a small group of senior peers to pressure-test your technical strategy. Starts April 15. Secure your spot: bit.ly/4tFcazC

Early in your career? The QCon San Francisco 2026 volunteer program gives you free access to sessions and recordings in exchange for onsite support. Conference: Nov 16-18, San Francisco. Apply: bit.ly/4sZ6sIX

AI infrastructure decisions are being shaped by constraints most teams can’t see from the inside. Jordan Nanos @SemiAnalysis breaks down the hardware and economics at QCon AI Boston, June 1-2, Boston. Early bird ends April 14. 🔗 bit.ly/4vfbBhu

Justifying architectural trade-offs to stakeholders requires more than technical intuition. The InfoQ Certified Architect cohort covers the Architecture Canvas, ADRs, and RFC frameworks. Starts April 15. Find out more: bit.ly/4vh0lkz

AI doesn't just speed up delivery; it exposes technical debt. DORA data from 5,000 professionals shows AI surfaces dysfunction faster than teams can pay it down. Get practitioner-led strategies at QCon AI Boston and QCon San Francisco. Early bird ends April 14: bit.ly/4dG7Ew9