That is the primary article in a sequence on agentic engineering and AI-driven growth. Search for the subsequent article on March 19 on O’Reilly Radar.
There’s been a whole lot of hype about AI and software program growth, and it is available in two flavors. One says, “We’re all doomed, that instruments like Claude Code will make software program engineering out of date inside a 12 months.” The opposite says, “Don’t fear, every little thing’s high-quality, AI is simply one other software within the toolbox.” Neither is sincere.
I’ve spent over 20 years writing about software program growth for practitioners, overlaying every little thing from coding and structure to venture administration and staff dynamics. For the final two years I’ve been targeted on AI, coaching builders to make use of these instruments successfully, writing about what works and what doesn’t in books, articles, and reviews. And I saved working into the identical downside: I had but to search out anybody with a coherent reply for the way skilled builders ought to really work with these instruments. There are many suggestions and loads of hype however little or no construction, and little or no you possibly can follow, train, critique, or enhance.
I’d been observing builders at work utilizing AI with numerous ranges of success, and I noticed we have to begin occupied with this as its personal self-discipline. Andrej Karpathy, the previous head of AI at Tesla and a founding member of OpenAI, just lately proposed the time period “agentic engineering” for disciplined growth with AI brokers, and others like Addy Osmani are getting on board. Osmani’s framing is that AI brokers deal with implementation however the human owns the structure, opinions each diff, and exams relentlessly. I believe that’s proper.
However I’ve spent a whole lot of the final two years instructing builders tips on how to use instruments like Claude Code, agent mode in Copilot, Cursor, and others, and what I maintain listening to is that they already know they need to be reviewing the AI’s output, sustaining the structure, writing exams, holding documentation present, and staying in command of the codebase. They know tips on how to do it in idea. However they get caught making an attempt to use it in follow: How do you really assessment 1000’s of traces of AI-generated code? How do you retain the structure coherent whenever you’re working throughout a number of AI instruments over weeks? How have you learnt when the AI is confidently unsuitable? And it’s not simply junior builders who’re having bother with agentic engineering. I’ve talked to senior engineers who battle with the shift to agentic instruments, and intermediate builders who take to it naturally. The distinction isn’t essentially the years of expertise; it’s whether or not they’ve discovered an efficient and structured option to work with AI coding instruments. That hole between realizing what builders ought to be doing with agentic engineering and realizing tips on how to combine it into their day-to-day work is an actual supply of hysteria for lots of engineers proper now. That’s the hole this sequence is making an attempt to fill.
Regardless of what a lot of the hype about agentic engineering is telling you, this type of growth doesn’t get rid of the necessity for developer experience; simply the alternative. Working successfully with AI brokers really raises the bar for what builders must know. I wrote about that have hole in an earlier O’Reilly Radar piece referred to as “The Cognitive Shortcut Paradox.” The builders who get essentially the most from working with AI coding instruments are those who already know what good software program seems to be like, and might typically inform if the AI wrote it.
The concept that AI instruments work finest when skilled builders are driving them matched every little thing I’d noticed. It rang true, and I needed to show it in a manner that different builders would perceive: by constructing software program. So I began constructing a selected, sensible method to agentic engineering constructed for builders to observe, after which I put it to the check. I used it to construct a manufacturing system from scratch, with the rule that AI would write all of the code. I wanted a venture that was complicated sufficient to stress-test the method, and fascinating sufficient to maintain me engaged by the onerous components. I needed to use every little thing I’d discovered and uncover what I nonetheless didn’t know. That’s once I got here again to Monte Carlo simulations.
The experiment
I’ve been obsessive about Monte Carlo simulations ever since I used to be a child. My dad’s an epidemiologist—his complete profession has been about discovering patterns in messy inhabitants knowledge, which suggests statistics was all the time a part of our lives (and it additionally signifies that I discovered SPSS at a really early age). After I was possibly 11 he informed me concerning the drunken sailor downside: A sailor leaves a bar on a pier, taking a random step towards the water or towards his ship every time. Does he fall in or make it dwelling? You possibly can’t know from any single run. However run the simulation a thousand instances, and the sample emerges from the noise. The person final result is random; the mixture is predictable.
I keep in mind writing that simulation in BASIC on my TRS-80 Coloration Laptop 2: a little bit blocky sailor stumbling throughout the display, two steps ahead, one step again. The drunken sailor is the “Hiya, world” of Monte Carlo simulations. Monte Carlo is a method for issues you may’t clear up analytically: You simulate them a whole lot or 1000’s of instances and measure the mixture outcomes. Every particular person run is random, however the statistics converge on the true reply because the pattern dimension grows. It’s a method we mannequin every little thing from nuclear physics to monetary threat to the unfold of illness throughout populations.
What when you might run that form of simulation at the moment by describing it in plain English? Not a toy demo however 1000’s of iterations with seeded randomness for reproducibility, the place the outputs get validated and the outcomes get aggregated into precise statistics you should utilize. Or a pipeline the place an LLM generates content material, a second LLM scores it, and something that doesn’t move will get despatched again for an additional strive.
The objective of my experiment was to construct that system, which I referred to as Octobatch. Proper now, the trade is continually in search of new real-world end-to-end case research in agentic engineering, and I needed Octobatch to be precisely that case research.
I took every little thing I’d discovered from instructing and observing builders working with AI, put it to the check by constructing an actual system from scratch, and turned the teachings right into a structured method to agentic engineering I’m calling AI-driven growth, or AIDD. That is the primary article in a sequence about what agentic engineering seems to be like in follow, what it calls for from the developer, and how one can apply it to your individual work.
The result’s a totally functioning, well-tested software that consists of about 21,000 traces of Python throughout a number of dozen recordsdata, backed by full specs, almost a thousand automated exams, and high quality integration and regression check suites. I used Claude Cowork to assessment all of the AI chats from all the venture, and it seems that I constructed all the software in roughly 75 hours of lively growth time over seven weeks. For comparability, I constructed Octobatch in simply over half the time I spent final 12 months taking part in Blue Prince.
However this sequence isn’t nearly Octobatch. I built-in AI instruments at each degree: Claude and Gemini collaborating on structure, Claude Code writing the implementation, LLMs producing the pipelines that run on the system they helped construct. This sequence is about what I discovered from that course of: the patterns that labored, the failures that taught me essentially the most, and the orchestration mindset that ties all of it collectively. Every article pulls a special lesson from the experiment, from validation structure to multi-LLM coordination to the values that saved the venture on observe.
Agentic engineering and AI-driven growth
When most individuals speak about utilizing AI to jot down code, they imply considered one of two issues: AI coding assistants like GitHub Copilot, Cursor, or Windsurf, which have developed nicely past autocomplete into agentic instruments that may run multifile enhancing classes and outline customized brokers; or “vibe coding,” the place you describe what you need in pure language and settle for no matter comes again. These coding assistants are genuinely spectacular, and vibe coding might be actually productive.
Utilizing these instruments successfully on an actual venture, nonetheless, sustaining architectural coherence throughout 1000’s of traces of AI-generated code, is a special downside completely. AIDD goals to assist clear up that downside. It’s a structured method to agentic engineering the place AI instruments drive substantial parts of the implementation, structure, and even venture administration, when you, the human within the loop, determine what will get constructed and whether or not it’s any good. By “construction,” I imply a set of practices builders can be taught and observe, a option to know whether or not the AI’s output is definitely good, and a option to keep on observe throughout the lifetime of a venture. If agentic engineering is the self-discipline, AIDD is one option to follow it.
In AI-driven growth, builders don’t simply settle for strategies or hope the output is appropriate. They assign particular roles to particular instruments: one LLM for structure planning, one other for code execution, a coding agent for implementation, and the human for imaginative and prescient, verification, and the selections that require understanding the entire system.
And the “pushed” half is literal. The AI is writing nearly all the code. One in all my floor guidelines for the Octobatch experiment was that I’d let AI write all of it. I’ve excessive code high quality requirements, and a part of the experiment was seeing whether or not AIDD might produce a system that meets them. The human decides what will get constructed, evaluates whether or not it’s proper, and maintains the constraints that maintain the system coherent.
Not everybody agrees on how a lot the developer wants to remain within the loop, and the totally autonomous finish of the spectrum is already producing cautionary tales. Nicholas Carlini at Anthropic just lately tasked 16 Claude situations to construct a C compiler in parallel with no human within the loop. After 2,000 classes and $20,000 in API prices, the brokers produced a 100,000-line compiler that may construct a Linux kernel however isn’t a drop-in alternative for something, and when all 16 brokers obtained caught on the identical bug, Carlini needed to step again in and partition the work himself. Even sturdy advocates of a totally hands-off, vibe-driven method to agentic engineering would possibly name {that a} step too far. The query is how a lot human judgment you might want to make that code reliable, and what particular practices assist you apply that judgment successfully.
The orchestration mindset
If you wish to get builders occupied with agentic engineering in the precise manner, it’s important to begin with how they give thought to working with AI, not simply what instruments they use. That’s the place I began once I started constructing a structured method, and it’s why I began with habits. I developed a framework for these referred to as the Sens-AI Framework, revealed as each an O’Reilly report (Essential Considering Habits for Coding with AI) and a Radar sequence. It’s constructed round 5 practices: offering context, doing analysis earlier than prompting, framing issues exactly, iterating intentionally on outputs, and making use of important considering to every little thing the AI produces. I began there as a result of habits are the way you lock in the way in which you consider the way you’re working. With out them, AI-driven growth produces plausible-looking code that falls aside beneath scrutiny. With them, it produces methods {that a} single developer couldn’t construct alone in the identical time-frame.
Habits are the inspiration, however they’re not the entire image. AIDD additionally has practices (concrete strategies like multi-LLM coordination, context file administration, and utilizing one mannequin to validate one other’s output) and values (the ideas behind these practices). In case you’ve labored with Agile methodologies like Scrum or XP, that construction ought to be fairly acquainted: Practices inform you tips on how to work day-to-day, and habits are the reflexes you develop in order that the practices grow to be computerized.
Values typically appear weirdly theoretical, however they’re an necessary piece of the puzzle as a result of they information your choices when the practices don’t provide you with a transparent reply. There’s an rising tradition round agentic engineering proper now, and the values you convey to your venture both match or conflict with that tradition. Understanding the place the values come from is what makes the practices stick. All of that results in an entire new mindset, what I’m calling the orchestration mindset. This sequence builds all 4 layers, utilizing Octobatch because the proving floor.
Octobatch was a deliberate experiment in AIDD. I designed the venture as a check case for all the method, to see what a disciplined AI-driven workflow might produce and the place it will break down, and I used it to use and enhance the practices and values to make them efficient and straightforward to undertake. And whether or not by intuition or coincidence, I picked the proper venture for this experiment. Octobatch is a batch orchestrator. It coordinates asynchronous jobs, manages state throughout failures, tracks dependencies between pipeline steps, and makes positive validated outcomes come out the opposite finish. That form of system is enjoyable to design however a whole lot of the main points, like state machines, retry logic, crash restoration, and value accounting, might be tedious to implement. It’s precisely the form of work the place AIDD ought to shine, as a result of the patterns are nicely understood however the implementation is repetitive and error-prone.
Orchestration—the work of coordinating a number of unbiased processes towards a coherent final result—developed right into a core thought behind AIDD. I discovered myself orchestrating LLMs the identical manner Octobatch orchestrates batch jobs: assigning roles, managing handoffs, validating outputs, recovering from failures. The system I used to be constructing and the method I used to be utilizing to construct it adopted the identical sample. I didn’t anticipate it once I began, however constructing a system that orchestrates AI seems to be a fairly good option to learn to orchestrate AI. That’s the unintentional a part of the unintentional orchestrator. That parallel runs by each article on this sequence.
Need Radar delivered straight to your inbox? Be a part of us on Substack. Join right here.
The trail to batch
I didn’t start the Octobatch venture by beginning with a full end-to-end Monte Carlo simulation. I began the place most individuals begin: typing prompts right into a chat interface. I used to be experimenting with totally different simulation and era concepts to offer the venture some construction, and some of them caught. A blackjack technique comparability turned out to be an incredible check case for a multistep Monte Carlo simulation. NPC dialogue era for a role-playing sport gave me a inventive workload with subjective high quality to measure. Each had the identical form: a set of structured inputs, every processed the identical manner. So I had Claude write a easy script to automate what I’d been doing by hand, and I used Gemini to double-check the work, make certain Claude actually understood my ask, and repair hallucinations. It labored high-quality at small scale, however as soon as I began working greater than 100 or so models, I saved hitting fee limits, the caps that suppliers placed on what number of API requests you may make per minute.
That’s what pushed me to LLM batch APIs. As a substitute of sending particular person prompts separately and ready for every response, the key LLM suppliers all supply batch APIs that allow you to submit a file containing all your requests directly. The supplier processes them on their very own schedule; you look ahead to outcomes as a substitute of getting them instantly, however you don’t have to fret about fee caps. I used to be blissful to find in addition they value 50% much less, and that’s once I began monitoring token utilization and prices in earnest. However the actual shock was that batch APIs carried out higher than real-time APIs at scale. As soon as pipelines obtained previous the 100- or 200-unit mark, batch began working considerably sooner than actual time. The supplier processes the entire batch in parallel on their infrastructure, so that you’re not bottlenecked by round-trip latency or fee caps anymore.
The change to batch APIs modified how I thought of the entire downside of coordinating LLM API calls at scale, and led to the concept of configurable pipelines. I might chain levels collectively: The output of 1 step might grow to be the enter to the subsequent, and I might kick off the entire pipeline and are available again to completed outcomes. It seems I wasn’t the one one making the shift to batch APIs. Between April 2024 and July 2025, OpenAI, Anthropic, and Google all launched batch APIs, converging on the identical pricing mannequin: 50% of the real-time fee in alternate for asynchronous processing.
You in all probability didn’t discover that every one three main AI suppliers launched batch APIs. The trade dialog was dominated by brokers, software use, MCP, and real-time reasoning. Batch APIs shipped with comparatively little fanfare, however they symbolize a real shift in how we are able to use LLMs. As a substitute of treating them as conversational companions or one-shot SaaS APIs, we are able to deal with them as processing infrastructure, nearer to a MapReduce job than a chatbot. You give them structured knowledge and a immediate template, and so they course of all of it and hand again the outcomes. What issues is that you could now run tens of 1000’s of those transformations reliably, at scale, with out managing fee limits or connection failures.
Why orchestration?
If batch APIs are so helpful, why can’t you simply write a for-loop that submits requests and collects outcomes? You possibly can, and for easy circumstances a fast script with a for-loop works high-quality. However when you begin working bigger workloads, the issues begin to pile up. Fixing these issues turned out to be one of the crucial necessary classes for creating a structured method to agentic engineering.
First, batch jobs are asynchronous. You submit a job, and outcomes come again hours later, so your script wants to trace what was submitted and ballot for completion. In case your script crashes within the center, you lose that state. Second, batch jobs can partially fail. Possibly 97% of your requests succeeded and three% didn’t. Your code wants to determine which 3% failed, extract them, and resubmit simply these gadgets. Third, when you’re constructing a multistage pipeline the place the output of 1 step feeds into the subsequent, you might want to observe dependencies between levels. And fourth, you want value accounting. Once you’re working tens of 1000’s of requests, you need to understand how a lot you spent, and ideally, how a lot you’re going to spend whenever you first begin the batch. Each considered one of these has a direct parallel to what you’re doing in agentic engineering: holding observe of the work a number of AI brokers are doing directly, coping with code failures and bugs, ensuring all the venture stays coherent when AI coding instruments are solely wanting on the one half presently in context, and stepping again to have a look at the broader venture administration image.
All of those issues are solvable, however they’re not issues you need to clear up again and again (in each conditions—whenever you’re orchestrating LLM batch jobs or orchestrating AI coding instruments). Fixing these issues within the code gave some fascinating classes concerning the general method to agentic engineering. Batch processing strikes the complexity from connection administration to state administration. Actual-time APIs are onerous due to fee limits and retries. Batch APIs are onerous as a result of it’s important to observe what’s in flight, what succeeded, what failed, and what’s subsequent.
Earlier than I began growth, I went in search of present instruments that dealt with this mixture of issues, as a result of I didn’t need to waste my time reinventing the wheel. I didn’t discover something that did the job I wanted. Workflow orchestrators like Apache Airflow and Dagster handle DAGs and job dependencies, however they assume duties are deterministic and don’t present LLM-specific options like immediate template rendering, schema-based output validation, or retry logic triggered by semantic high quality checks. LLM frameworks like LangChain and LlamaIndex are designed round real-time inference chains and agent loops—they don’t handle asynchronous batch job lifecycles, persist state throughout course of crashes, or deal with partial failure restoration on the chunk degree. And the batch API consumer libraries from the suppliers themselves deal with submission and retrieval for a single batch, however not multistage pipelines, cross-step validation, or provider-agnostic execution.
Nothing I discovered lined the complete lifecycle of multiphase LLM batch workflows, from submission and polling by validation, retry, value monitoring, and crash restoration, throughout all three main AI suppliers. That’s what I constructed.
Classes from the experiment
The objective of this text, as the primary one in my sequence on agentic engineering and AI-driven growth, is to put out the speculation and construction of the Octobatch experiment. The remainder of the sequence goes deep on the teachings I discovered from it: the validation structure, multi-LLM coordination, the practices and values that emerged from the work, and the orchestration mindset that ties all of it collectively. A couple of early classes stand out, as a result of they illustrate what AIDD seems to be like in follow and why developer expertise issues greater than ever.
- It’s a must to run issues and examine the information. Bear in mind the drunken sailor, the “Hiya, world” of Monte Carlo simulations? At one level I observed that once I ran the simulation by Octobatch, 77.5% of the sailors fell within the water. The outcomes for a random stroll ought to be 50/50, so clearly one thing was badly unsuitable. It turned out the random quantity generator was being re-seeded at each iteration with sequential seed values, which created correlation bias between runs. I didn’t establish the issue instantly; I ran a bunch of exams utilizing Claude Code as a check runner to generate every check, run it, and log the outcomes; Gemini appeared on the outcomes and located the foundation trigger. Claude had bother arising with a repair that labored nicely, and proposed a workaround with a big checklist of preseeded random quantity values within the pipeline. Gemini proposed a hash-based repair reviewing my conversations with Claude, but it surely appeared overly complicated. As soon as I understood the issue and rejected their proposed options, I made a decision the very best repair was easier than both of the AI’s strategies: a persistent RNG per simulation unit that superior naturally by its sequence. I wanted to know each the statistics and the code to guage these three choices. Believable-looking output and proper output aren’t the identical factor, and also you want sufficient experience to inform the distinction. (We’ll discuss extra about this case within the subsequent article within the sequence.)
- LLMs typically overestimate complexity. At one level I needed so as to add help for customized mathematical expressions within the evaluation pipeline. Each Claude and Gemini pushed again, telling me, “That is scope creep for v1.0” and “Reserve it for v1.1.” Claude estimated three hours to implement. As a result of I knew the codebase, I knew we had been already utilizing asteval, a Python library that gives a secure, minimalistic evaluator for mathematical expressions and easy Python statements, elsewhere to guage expressions, so this appeared like a simple use of a library we’re already utilizing elsewhere. Each LLMs thought the answer can be way more complicated and time-consuming than it really was; it took simply two prompts to Claude Code (generated by Claude), and about 5 minutes complete to implement. The function shipped and made the software considerably extra highly effective. The AIs had been being conservative as a result of they didn’t have my context concerning the system’s structure. Expertise informed me the combination can be trivial. With out that have, I’d have listened to them and deferred a function that took 5 minutes.
- AI is commonly biased towards including code, not deleting it. Generative AI is, unsurprisingly, biased towards era. So once I requested the LLMs to repair issues, their first response was typically so as to add extra code, including one other layer or one other particular case. I can’t consider a single time in the entire venture when one of many AIs stepped again and mentioned, “Tear this out and rethink the method.” The best classes had been those the place I overrode that intuition and pushed for simplicity. That is one thing skilled builders be taught over a profession: Probably the most profitable modifications typically delete greater than they add—the PRs we brag about are those that delete 1000’s of traces of code.
- The structure emerged from failure. The AI instruments and I didn’t design Octobatch’s core structure up entrance. Our first try was a Python script with in-memory state and a whole lot of hope. It labored for small batches however fell aside at scale: A community hiccup meant restarting from scratch, a malformed response required guide triage. Quite a lot of issues fell into place after I added the constraint that the system should survive being killed at any second. That single requirement led to the tick mannequin (get up, examine state, do work, persist, exit), the manifest file as supply of reality, and all the crash-recovery structure. We found the design by repeatedly failing to do one thing easier.
- Your growth historical past is a dataset. I simply informed you many tales from the Octobatch venture, and this sequence shall be stuffed with them. Each a type of tales got here from going again by the chat logs between me, Claude, and Gemini. With AIDD, you might have a whole transcript of each architectural choice, each unsuitable flip, each second the place you overruled the AI and each second the place it corrected you. Only a few growth groups have ever had that degree of constancy of their venture historical past. Mining these logs for classes discovered seems to be one of the crucial useful practices I’ve discovered.
Close to the top of the venture, I switched to Cursor to verify none of this was particular to Claude Code. I created recent conversations utilizing the identical context recordsdata I’d been sustaining all through growth, and was in a position to bootstrap productive classes instantly; the context recordsdata labored precisely as designed. The practices I’d developed transferred cleanly to a special software. The worth of this method comes from the habits, the context administration, and the engineering judgment you convey to the dialog, not from any specific vendor.
These instruments are shifting the world in a path that favors builders who perceive the methods engineering can go unsuitable and know strong design and structure patterns…and who’re okay letting go of management of each line of code.
What’s subsequent
Agentic engineering wants construction, and construction wants a concrete instance to make it actual. The subsequent article on this sequence goes into Octobatch itself, as a result of the way in which it orchestrates AI is a remarkably shut parallel to what AIDD asks builders to do. Octobatch assigns roles to totally different processing steps, manages handoffs between them, validates their outputs, and recovers once they fail. That’s the identical sample I adopted when constructing it: assigning roles to Claude and Gemini, managing handoffs between them, validating their outputs, and recovering once they went down the unsuitable path. Understanding how the system works seems to be a great way to know tips on how to orchestrate AI-driven growth. I’ll stroll by the structure, present what an actual pipeline seems to be like from immediate to outcomes, current the information from a 300-hand blackjack Monte Carlo simulation that places all of those concepts to the check, and use all of that to reveal concepts we are able to apply on to agentic engineering and AI-driven growth.
Later articles go deeper into the practices and concepts I discovered from this experiment that make AI-driven growth work: how I coordinated a number of AI fashions with out dropping management of the structure, what occurred once I examined the code towards what I really meant to construct, and what I discovered concerning the hole between code that runs and code that does what you meant. Alongside the way in which, the experiment produced some findings about how totally different AI fashions see code that I didn’t count on—and that turned out to matter greater than I assumed they’d.
