Jellyfish

The leading intelligence platform for AI-Integrated engineering, helping more than 1,000 companies leverage AI to transform how they build software.

[Jellyfish @_jellyfish_co]

What's one AI-era engineering belief you think most leaders have completely wrong? 🤔

We asked these folks and got their hot takes:
- Garrett Moedl (Head of Product Marketing at Adopt AI)
- Latha Rao (Director of Product Management at Intel Corporation)
- Stephan Donze (CEO at AO Docs)
- Prathusha Prakash (Engineering Manager at TikTok)
- Aaron Sua (Sr Director of Platform Engineering at American Red Cross)

#AIEngineering #HotTakes

[Jellyfish @_jellyfish_co]

"What we see a lot is that people are just kicking the cans down the road. 'Look at how many pull requests I have. This is amazing.' The token maxing thing. 'Look at how many tokens I used.' And then you realize that, 'okay, and now it's stalled because the code's not passing your test or it's coming up with a lot of sonar issues or it's actually breaking in reality. So you have to hit pause on stuff right in production.' So I think you're shifting bottlenecks and the teams that work the best are the ones that really think about how do you do this guide, verify, solve thing as part of your process."

Watch Sonar CEO Tariq Shaukat's full ✨ GLOW episode here: https://t.co/NfSDRrhZMS

#GLOW26 #AIFirst #AIEngineering

[Jellyfish @_jellyfish_co]

Concurrency hits a ceiling: 84% run just one or two agents at once 🤖

If running agents for long periods of time in a loop with a goal until it completes is what unlocks output, the obvious next step is to run more agents at once. But we discovered that high concurrency levels are rare.

Measuring peak concurrent agent use across customers’ turns (each session is composed of multiple turns) hit a ceiling: 84% of active users top out at one or two agents at a time, and even the small group launching four or more spends over 80% of session time attending to a single agent.

This is the agentic barrier at its clearest. Human attention is the blocker on multi-agent workflows, and attention does not parallelize. Scaling infrastructure to orchestrate more than a pair of agents is not trivial either, and the industry is still building the tooling for users to scale agents.

Continue reading the State of AI report here: https://t.co/XTJS419HqF

#StateOfAI #AIEngineering #AITrends #AgenticBarrier

[Jellyfish @_jellyfish_co]

"As a customer told me, if any of my teams delivered 80% accuracy, I'd fire them. Right? And so you start getting to, all right, what about 90%, what about 95%? And I think the best teams are the ones who actually realize that it's not accurate and realize that it needs to be managed and those teams actually, in my experience, anecdotally move faster than the teams that actually trust in the AI too much, the cognitive surrender thing."

✨ Watch Sonar CEO Tariq Shaukat's full take on who's verifying all the code AI is writing here: https://t.co/NfSDRrhZMS

#GLOW26 #AIEngineering #AIFirst

[Jellyfish @_jellyfish_co]

We're hiring for multiple open positions, including these roles on our R&D teams:

📙 Senior Backend Software Engineer
We're seeking a Senior Backend Engineer who's passionate about building robust data pipelines, integrating with third-party APIs, and working with large-scale data infrastructure. Ideally, you're performance-minded and have experience with modern ETL patterns, workflow orchestration engines, and building API clients. This is a great opportunity to be a member of an exceptional engineering organization with high agency: where you can help shape the features your team builds and how your team builds them.

📕 Staff Data Engineer
Working closely with our Lead Data Architect, you’ll be responsible for implementing core data models, building production-grade CI/CD for data pipelines, and transforming raw engineering signals into highly optimized analytical layers. If you view broken pipelines and manual data patches as a technical debt to be solved and want to write code that directly impacts how the world’s best engineering leaders measure their output, you’re the perfect fit.

📘 Senior Data and AI Platform Engineer
This role is ideal for a senior individual contributor who is excited to combine traditional data engineering with modern data platform ownership. You will help build the data and AI platform that powers our internal enterprise-wide analytics, self-service BI, agentic analytics, and AI-enabled workflows. You will still build ingestion, transformation, and data delivery pipelines, but your broader mandate will be to help make Databricks the trusted foundation for analytics and AI across the company.

See all open jobs here and apply today: https://t.co/tfo9owt7nu

#Careers #JobOpenings #SoftwareJobs

[Jellyfish @_jellyfish_co]

🎉 The newest episode of ✨ GLOW "Leading Voices at the Intersection of AI and Engineering" is here...featuring Tariq Shaukat, CEO at Sonar.

In this 23-minute episode, hear Tariq's takes on:
- what happens when organizations move from code generation to code verification
- the rise of "cognitive surrender"
- why AI-generated code is often plausible but not necessarily correct
- how engineering leaders can balance productivity, quality, security, and accountability in the age of agentic development

Watch now: https://t.co/NfSDRrhZMS

[Jellyfish @_jellyfish_co]

🔎 Get an exclusive look at new data from the first half of 2026 as Jellyfish Research explore why some engineering organizations are turning exploding AI spend into measurable business outcomes while others are simply burning through tokens.

In this live research briefing on August 5, Tomas Pardinas (Sr Product Researcher) and Ben Kotrc (Research Director) explore:
- Why AI spending is growing faster than expected (and what to do about it)
- Understanding effective AI costs and token economics
- How context, workflows, and codebase design outperform simply buying more AI
- What engineering leaders should prioritize during the second half of 2026

Register today: https://t.co/B5JKHsSpYi

#StateOfAI #AIEngineering #AITrends

[Jellyfish @_jellyfish_co]

"We have a baseline of before AI came into play. We can really see the inflection point – the lifecycle of an issue from beginning to end started to shorten. In our case, 30-40% is what we’ve already seen, and our best teams are two to three times faster."

Learn more about how LastPass CTO Jason Rasmussen and team accomplished this here: https://t.co/lU7Lo8Naf6

#CaseStudy #CustomerStory #AIEngineering

[Jellyfish @_jellyfish_co]

"AI can't read your mind. I wish it did, but it's generally eager and generally unwilling to tell you it doesn't know the answer. If your peer wouldn't understand what you're asking, AI won't either."

Catch Joe Cudby's (Principal Product Manager - Agentic AI at AWS) insights and AI hot takes on in his 11-minute ✨ GLOW episode here: https://t.co/NfSDRrhZMS

#GLOW26 #AIFirst #AIEngineering

[Jellyfish @_jellyfish_co]

Flo Health CTO Roman Bugaev shares that "our investors like the feature where we can benchmark ourselves to other companies. It answers a lot of questions from investors like whether we are adopting AI with the speed expected from the market, whether we have proper performance management, and how we allocate resources. It gives a lot of tools for us to speak with investors in a language that they understand."

Learn more here: https://t.co/XS6oU4oCxk

#CustomerStories #CaseStudy #AIStory

[Jellyfish @_jellyfish_co]

"Structured enablement matters. It's not enough to provide the tool. Organizations that invest in deliberate onboarding, clear use cases, and ongoing support, see meaningfully better outcomes than those that just flip the switch and hope. But the fluency? The fluency comes from using AI in your environment."

Get Joe Cudby's (Principal Product Manager - Agentic AI at AWS) full take on where the biggest gains from AI come from in his 11-minute ✨ GLOW episode here: https://t.co/NfSDRrhZMS

#GLOW26 #AIFirst #AIEngineering

[Jellyfish @_jellyfish_co]

🧪 Jellyfish Research has been tracking token usage, and in January something changed. November 2025 brought the release of Opus 4.5, Anthropic’s main model, and usage patterns shifted sharply afterward.

Token consumption climbed steeply from the start of 2026, tripling for power users (P90) from 50M to 170M. 📈

We can also see a divergence between power users (P90) and the median (P50). The first group completely decoupled from the rest of the group, benefiting from the increase in throughput and widening the gap from a 50M token difference to 150M tokens.

👉 The question becomes: who is justifying the bill, and why most cannot follow?

If a thin tail is pulling away, the obvious question is what that tail does differently. Three frontier behaviors separate them: they put more work through agents, they let agents run longer with less supervision, and they push on running more agents at once. Each one helps explain who sits on the paying-off side of the divergence, and why the rest are locked out for now.

Dive into the data when you download the State of AI for Software Engineering (H1 '26) report here: https://t.co/YscaCzU9F1

#StateOfAI #AIEngineering #AIUsage #TokenSpend

[Jellyfish @_jellyfish_co]

To drive a successful AI transformation, you need 2️⃣ things: the right adoption playbook and the metrics to prove it’s working.

🤝 Together, DevClarity and Jellyfish help engineering organizations successfully adopt AI tools and prove the business value of every dollar spent:

- DevClarity enables your engineering teams to adopt and implement AI across the SDLC, directly improving their day-to-day capacity to build and ship high-quality software.

- Jellyfish provides the telemetry, measuring which teams are utilizing AI most consistently and identifying the cross-team process improvements needed to unlock full organizational velocity.

Ready to adopt and leverage AI in a meaningful way? Start here: https://t.co/O6OmCB5jmx

[Jellyfish @_jellyfish_co]

🎉 Just released: The C in SPACE: What AI Productivity Metrics Miss by AWS Principal Product Manager - Agentic AI, Joe Cudby.

In this 11-minute episode, hear Joe's takes on:

- why the biggest gains from AI come from something much harder to measure – communication and collaboration

- how knowledge sharing, living documentation, organizational fluency, and team dynamics shape AI success

- why the future of engineering productivity may depend less on the tools themselves and more on how people work together

Watch now: https://t.co/NfSDRrhZMS

#GLOW26 #AIFirst #AIEngineering

[Jellyfish @_jellyfish_co]

What is the most misleading AI metric that engineering leaders are using right now? 🤔

We asked these folks for their hot takes, and they delivered:
- Stephan Donze (CEO at AO Docs)
- Prathusha Prakash (Engineering Manager at TikTok)
- Aaron Sua (Sr Director of Platform Engineering at American Red Cross)
- Garrett Moedl(Head of Product Marketing at Adopt AI)
- Sedky Haider (Field CTO at Tyk)

#AIHotTakes #AIEngineering #HotTakes

[Jellyfish @_jellyfish_co]

"Now, I think this is a really exciting time for us as platform engineers. The best leverage that we can provide is enabling the business to go solve the business problems. So I think we have an opportunity, a really exciting opportunity as leaders in this space, to think about how do we leverage AI to enable those citizen developers and those builders all the way across our organization."

Get a front row seat to Google Cloud's DORA Lead Nathen Harvey's thoughts on how platform teams can create the foundations that help humans and AI work together to solve business problems faster in his ✨ GLOW episode here: https://t.co/NfSDRrhZMS

#GLOW26 #AIFirst #AIEngineering

[Jellyfish @_jellyfish_co]

In case you missed it, ✨ GLOW episodes are in full swing:

- Loadsmart CTO Ron Ben Yosef: In an AI-Driven Codebase, Engineers Matter More Than Ever

- https://t.co/cNMAMrwg5n Co-Founder and Field CTO Chris Evans: The Bottleneck isn’t Building - Operating Systems in the Age of AI

- Avaya VP of Engineering Ruby Agarwal: From AI Demo to Production Reality

- Q2 VP, Engineering Operations Jan Acosta: AI is Rewriting Software Delivery

- Augment Code VP of Engineering Vinay Perneti: Rethinking Code Review, Hiring, and Org Design in the Agent Era

- Flo Health Director of Engineering for AI Platform Vladislav Nedosekin: From Code to Care: Building Safe, Useful AI for 500M+ Users

- OpenHands Co-Founder & CEO Robert Brennan: Inner Loop vs. Outer Loop:
Where AI Agents Deliver the Most Value in Software Development

- Google Cloud DORA Lead Nathen Harvey: Building for Humans, Agents, and Everyone In Between

Stay tuned for many more episodes to come!
Watch now and get notified when new ones drop: https://t.co/NfSDRrhZMS

#GLOW26 #AIFirst #AIEngineering