Recorded at #IBMTechXchange 2026 in Atlanta, hear @KateHolterhoff chat with @myfear, AI Developer Strategist at @IBM, about the keynote announcements around Bob 2.0, including Bob Desktop, Bob Bazaar, and Bob DevSecOps. Full conversation on the MonkCast redmonk.com/videos/markus-eiโฆ
Welcome back to Tech Term of the Week! ๐ฅณ
This week's term โ agentic workflows - /ษหdสษn.tษชk หwษหk.floสz/
Definition โ AI-driven processes where autonomous AI agents make decisions, take actions and coordinate tasks with minimal human intervention.
Why it matters โ this allows AI systems to adapt and complete tasks even when conditions change, reducing the need for human intervention on routine work while also generating higher-quality outputs that can help train future AI models more effectively.
IBM Developer retweeted
Thatโs a wrap on #Current26 San Francisco! ๐
Hereโs to the people, the energy, and the moments that made this week unforgettable.
IBM Developer retweeted
Spend less time troubleshooting pipeline breaks and painful cluster migrations and more time building your Streamhouse.
Check out the new updates that make Confluent Cloud and Confluent Cloud simpler to build on, simpler to run and more cost-effective: ibm.co/6017EV0CH
Today, in collaboration with @IBMResearch, we are launching ScarfBench (Self-Contained Application Refactoring Benchmark), an open benchmark designed to test whether coding agents can successfully migrate enterprise Java apps across frameworks like Jakarta EE, Quarkus, and Spring while maintaining the appโs expected behavior.
Moving between frameworks takes coordinated changes across dependency injection, persistence, build systems, and deployment infrastructure. An agent can easily generate a migration that looks fine or compiles, only for the app to fail when deployed. ScarfBench puts that to the test.
ScarfBench evaluates each model with its harness over 204 migration tasks. Gemini 4 Argon with Antigravity scores 75.2%, followed by Claude Opus 5.5 with Claude Code at 70.5%, and Claude Opus 5 with Claude Code at 69.8%.
IBM Bob Self-Hosted is now supporting @nvidia Nemotron and Poolside Laguna.
After a year of rigorous model evaluation and optimization alongside NVIDIA, the IBM Bob team selected these models to power on-premises and air-gapped enterprise deployments.
Read more about the selection process โ ibm.co/6017Epm5d
IBM Developer retweeted
Bob TV: Episode 2: with yours truly and @if54uran !
Tokenmaxxing: How Much Context Does AI Coding Really Need?
Tue, Oct 6, 2026 11:00 AM EDT (5:00 PM Europe/Berlin)
ibm.webcasts.com/starthere.jโฆ
#Register #Free @IBMDeveloper #IBM
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
๐ developers ๐
Happy world developer day :)
IBM Developer retweeted
Canโt miss me being welcomed by Ryan to the @StackOverflow #podcast to talk about why your coding agent should be writing #Java. We talk about why the long history of Java both makes for a stable language and a source of training data for AI. On top of that, we discuss how agentic Java excels by working with powerful partners: a massive ecosystem of Java libraries and agentic harnesses.
stackoverflow.blog/2026/09/0โฆ
@IBM @IBMDeveloper
1 day left to submit to Agentic Cinema: The Blockbuster Hackathon โฐ Don't wait until the last minute. Start on your submissions on Devpost today! Check out the full submission checklist: bit.ly/agenticcinema26t
Where are you in your build? ๐
IBM Developer retweeted
Decades ago we were talking about SOA. Are we soon to be talking about #AOA: Agent Oriented Architecture?
Here's a sample application combining two agent harnesses into a multi-harness concept, using @GitHubCopilot and @IBM's Bob.
// @if54uran @myfear @Chels_LA @IBMDeveloper @ryanj @robdthomas
github.com/brunoborges/bob-cโฆ
IBM Developer retweeted
I am almost ready to travel to Berlin this week for @WeAreDevs ! Can't wait to see y'all there. Don't miss my session on Friday!
Code is cheap now. Software still isnโt. This talk is about using AI as leverage, not as a shortcut.
- AI can write Java code fast. Iโll show where that helps, where it fails, and how to stay in control in real systems.
- Most โhallucinationsโ are not magic. They are missing context. Iโll explain how to fix that.
- If you work on large Java codebases, Iโll share practical patterns for using AI without breaking what already works.
- Weโll look at constraints, decomposition, rules, skills, and verification, and why they matter more than prompt tricks.
Early enough to catch before you head home. Late enough so you don't get to wake up too early!
Right in the sweet spot! Also make sure to stop by the IBM Booth and grab some Bob swag while we have it! And of course try out Bob yourself!
See you all tomorrow!
IBM Developer retweeted
Iโll be talking to @KateHolterhoff and you can watch us chat about Agentic AI and share thoughts around how software teams work in this brave new world.
Register. Itโs free.
23 June 2026 at 13:30 GMT-4
#RedMonk #IBM
ibm.com/forms/mkt-webinar-90โฆ
@IBM
IBM Developer retweeted
everyone's building simple agents
meanwhile IBM is building robust enterprise agents in production, and it's open-source
they just dropped a blog on HF breaking down how to go beyond LLMs & agents: structured reasoning, tool use, and more to scale AI across enterprise
IBM Developer retweeted
Artificial Analysis and IBM Research are launching ITBench-AA, the first in a new series of benchmarks evaluating models on agentic enterprise IT tasks, starting with Site Reliability Engineering tasks where frontier models score below 50%
ITBench-AAโs SRE tasks benchmark model performance on Kubernetes incident response, where models must diagnose live systems by reading logs, tracing dependencies, and identifying root-cause entities across complex infrastructure. The underlying ITBench dataset has been developed by @IBM's Software Innovation Lab, leveraging IBMโs deep expertise in enterprise IT operations
Artificial Analysis has worked closely with IBM over the last 6 months to develop a implementation of the dataset for frontier AI evaluation, beginning with Site Reliability Engineering (SRE) and expanding to Financial Operations (FinOps) and Chief Information Security Officer (CISO) tasks over time
ITBench-AA SRE overview:
โค 59 SRE tasks in total: 40 public tasks and 19 brand new, held-out tasks
โค Each task provides a Kubernetes incident snapshot containing alerts, events, traces, metrics, logs, and application topology. The model must identify the minimal set of independent root-cause Kubernetes entities responsible for the incident
โค Faults span typical SRE failure modes including infrastructure, service, application, and chaos-injected incidents, such as resource quota exhaustion, rollout failures, connection pool exhaustion, and network partitions
Methodology details:
โค Agentic harness: each task is solved by the model running in our open-source Stirrup reference harness, with shell access to a sandboxed file system containing the relevant logs and snapshots. 100-turn cap per task, 3 repeats per task
โค Models submit a list of root-cause entities (Kubernetes Deployments, Services, Pods, etc.) they believe caused the incident. Each submission is compared against a ground-truth set of root causes provided by IBM Research
โค Scoring uses average precision at full recall: if a model misses any of the ground-truth root causes, it scores 0.0 for that repeat. If it identifies all of them, it is awarded a score equal to its precision - the share of its submitted entities that are actual root causes, i.e. true positives / (true positives + false positives). The headline score is the average across 59 tasks ร 3 repeats.
โค The harness (Stirrup) is held constant across all evaluated models, allowing an apples-to-apples comparison between models.
Key findings:
โค Claude Opus 4.7 (Adaptive Reasoning, Max Effort) leads at 47%, followed by GPT-5.5 (xhigh) at 46% and Qwen3.7 Max at 42%
โค All frontier models score below 50%, making ITBench-AA SRE one of the least saturated agentic benchmarks in our suite. For context, frontier models score considerably higher on Terminal-Bench
โค Turn counts vary nearly 3x and longer trajectories do not translate to higher accuracy. GPT-5.5 (xhigh) averages 31 turns per task at 46%, while Gemini 3.1 Pro Preview averages 83 turns at 30%. Models that over-investigate tend to surface upstream fault-injection mechanisms or co-occurring symptoms as false positives
โค GLM-5.1 (Reasoning) leads open weights models at 40%, effectively tied with Gemini 3.5 Flash (high). DeepSeek V4 Pro (Reasoning, Max Effort) follows at 38%, with Gemma 4 31B (Reasoning) at 37%, ahead of Gemini 3.1 Pro Preview at 30%
IBM Developer retweeted
Expose a Legacy Java App as an MCP Server Using a Kubernetes Sidecar Expose a Legacy Java App as an MCP Server Using a Kubernetes Sidecar:
๐๏ธ: No changes on the legacy app
๐ณ๏ธ: Using Containers and Kubernetes
โป๏ธ: No side effects
youtube.com/watch?v=3llzsh4Uโฆ
IBM Developer retweeted
I get suspicious when trim().toUpperCase() turns into an agent.
This walkthrough shows where MCP tools fit in a LangChain4j workflow: one Quarkus MCP server, one workflow app, /topology, and tests against Ollama.
buff.ly/OiFml9q
#langchain4j #java #quarkus #ollama



























