jaynate

>> “As engineering capacity is no longer a bottleneck, some engineering teams are starting to experience product starvation”

This has always been a problem that Marty Cagan calls “feeding the beast”, but it’s much worse now. Do engineers need to become more cross-functional themselves as well as their product mgt, UX, and quality control counterparts?

> “Then you tried to ship real client work, and the dream fell apart.”

For me it was here. I have tried for months to get AI to write for me and have now fully embraced the fact that LLMs are not ready to do that for me. None of them. I’ve gone back to writing my own content and only using the LLM to brainstorm.

The underpants gnomes. Well played!

1000%. The expense of writing software was and still is high.

Whether it’s commercial software company or an internal team writing custom software, the return on that investment depends on many things outside of the code itself.

When I initially saw this I was excited because what I’m trying to solve for is visibility of both semi and fully autonomous and agents at runtime. Marble is an improvement over chat and Claude but it’s incremental IMO.

We’ve toyed with a meta agent style interface which you chat to do everything. Add tools, connect datasets, build agents, deploy and schedule them, build skills, etc.

But at some point agents have to have an interface for human approvals, validation, training, and tuning.

I guess it’s like a spreadsheet. Ultimately agents are part of custom apps which are each going to be as unique like snowflakes. So it’s difficult to have one standard abstracted runtime interface without feeling obtuse to a business user. I think what I’m describing and wanting are new app layer building blocks vs one interface to rule them all.

All of this said, I’m all for a better personal llm harness interface (which is what this is). So keep up the great work guys!

>> Thanks for reading! Let me know if anyone's interested in more posts about this bowling nonsense.

I’d say, “yes!” judging from the upvotes.

Are you running in any enterprise environments that you know of?

Wow, can’t understand why this isn’t getting more attention.

For clarity, inference is typically a COGS and therefore hits Gross Margin vs model training which would typically be in OpEx (where R&D lives) and would hit operating margin.

Yes, and the free-for-all building of nonsense (and insecure) apps by non engineers is probably going to slow down as well.

Feels like we’re continuing to trend toward deterministic workflows which may actually be okay in 90% of cases. Reality is there’s a lot of unnecessary token burn happening right now. Simple market dynamics will solve that, i.e., when token cost subsidies begin to fade away and we face the true cost of agent applications.

Simply... amazing

Isn’t there coverage on any other site with fewer ads and popups? I could literally barely navigate the article on my phone.

And I think it’s less about letting agents modify the product source. That’s more of a platform capability which should also be a requirement for certain types of use cases. All comes back to listening to and / or innovating for customers.

“Their relationship with the software is one of pure dependency, and when the software doesn’t do what they need, they just… live with it”

Or, more likely, they churn off the product.

The SaaS platforms that will survive are busy RIGHT NOW revamping their APIs, implementing oauth, and generally reorganizing their products to be discovered and manipulated by agents. Failing in this effort will ultimately result in the demise of any given platform. This goes for larger SaaS companies, too, it’ll just take longer.

Good engineers are way more important than they’ve ever been and the job market tells the story. Engineering job posts are up 10% year over year. The work is changing but that’s what happens when a new technology wave comes ashore. Don’t give up, ride the new wave. You’re uniquely qualified.

I think about this book often since Nov 2022. It’s a fantastic read.

> The downside is that the dryRun-flag pollutes the code a bit. In all the major phases, I need to check if the flag is set, and only print the action that will be taken, but not actually doing it.

Sounds like a case for the state machine pattern

I think we are both saying similar things here (believe it or not). Sales leaders turnover with surprising frequency - 18-24 months. About the time the sales team tells you what they “want” and you fine tune it, they will be gone. The next person will come in and probably scrap 50-75% of what the prior leader did. New requirements.

Meanwhile, besides functionality, you’ll want/need to plug in the latest and greatest go to market tools for marketing -and demand gen. But… that’ll be a custom effort, too.

Along the way you’ll also realize that you’re missing out on the most common practices in the industry because you built some idiosyncratic tool that only is relevant to your company.

History may very well prove me wrong, but I think you’re underestimating the expertise that underlies these products and platforms. It’s not just code, and the costs of getting it wrong are more than just an engineer’s time. When you waste time in GTM the impacts on the business and valuation are not linear, they’re exponential.

>> Our sales people can make feature requests

I can tell you with near-100% certainty. This isn’t what you want to happen. Disaster in the making.

Just because you can doesn’t mean you should. There is very little competitive advantage to be gained from this type of effort in most companies.

I wish I could upvote this more than once. The author gets it, you have to sell outcomes. Not features. Seems like every open source company that doesn’t market an outcome to buyers will face a similar threat. And this particular go to market strategy was “brittle” before AI.

It’s sort of difficult to understand why this is even a question - LLM-based / judgment dependent workflows vs script-based / deterministic workflows.

In mapping out the problems that need to be solved with internal workflows, it’s wise to clarify where probabilistic judgments are helpful / required vs. not upfront. If the process is fixed and requires determinism why not just write scripts (code-gen’ed, of course).

Many of these laws can be applied to any discipline. Business, politics, etc.

My gmail address is first.last@gmail.com. From time to time (and for years) I get someone else’s at firstlast@gmail.com. I thought that a Gmail account that was first.last@gmail also allowed for email sent to firstlast@gmail (no period) to reach my inbox as well.

I’ve received some sensitive/PII content over the years.

I’ve wondered if this person has access to any of my information?

Not necessarily related to this post, but wonder why and how this could happen.

Try Pendo instead…

There are revenue recognition rules that govern what Revenue is on the P&L. Same with costs/expenses. But I would say if anything, profits are easier to manipulate than revenue.

Depending on what you’re optimizing for could be revenue (high growth) or Net income per share (lower growth).

Reality is once you’ve established a baseline it’s difficult to move from one to the other or have substantial changes to the negative for either.

Had me hooked right up to this point: “When we get the check, we pay a commission to this representative. Assume the commission is $500.”

I’ve never seen a software company pay 50% commissions on a software sale. I know it’s and example but the percentages are wrong even for the perpetually licensed days. Should be closer to 8-15%.

Totally sales and marketing spend could indeed be higher in this model because autodesk moved to direct positioning with end buyers rather than distributors.

“The transaction positions EA to accelerate innovation and growth to build the future of entertainment.”

Unless some of that cash is staying on the balance sheet, they will NOT be investing in growth and innovation.

PE is not what I normally think of when I think of innovation and (non-financially engineered) growth.