# JCody Ventures Good ideas deserve the ability to grow. - Site: https://jcody.com/ ## Frederick Brooks told you this would happen. In 1971. - Link: https://jcody.com/insights/frederick-brooks-mythical-man-month-ai-agents/ - Published: 2026-10-26T19:30:05-04:00 - Author: Jeffrey Cody Jake Van Clief ([@lostandlucky](https://www.instagram.com/lostandlucky/)) posted a reel this week about _The Mythical Man-Month_. Worth a watch. Here’s what I’d build on top of it — because there’s a piece that usually gets skipped. [Watch the reel →](https://www.instagram.com/reel/De9cW_Zxc_t) ## What Brooks actually said Frederick Brooks published _The Mythical Man-Month_ in 1975 (the research goes back to the early ’70s). The headline finding: adding more people to a late software project makes it later. Not because people are bad at their jobs. Because coordination overhead grows faster than output. Every new team member doesn’t just add capacity— they add communication paths. The work multiplies. The throughput doesn’t keep up. Companies are running straight into this problem with AI agents right now. Stand up a swarm — one agent to research, one to write, one to review, one to route — without a clear structure underneath it, and you get faster confusion. Not faster results. ## The piece that gets skipped Brooks wasn’t just writing about headcount. He was writing about structure. The fix for a late, overcrowded project wasn’t fewer people — it was clearer roles, cleaner handoffs, and someone who actually owned the architecture of the work itself. That’s where I see teams get stuck with AI. They buy the tooling, spin up the agents, and then hit a wall — because nobody built the structure the agents are supposed to operate inside. The agents are capable. The map doesn’t exist yet. No amount of prompting fixes a missing map. ## What getting it right early actually looks like Before you scale, answer these: What is each agent responsible for? What’s the source of truth when two agents surface conflicting information? Where does a human need to be in the loop, and why? How does work get handed off cleanly so nothing falls between the cracks? These aren’t AI questions. They’re operations questions. They’re the same questions good information architects and workflow designers have been asking for decades. The tooling is new. The discipline isn’t. Getting these sorted before you build saves you from rebuilding later — and rebuilding later, when agents are already embedded in real processes, is expensive and slow. Brooks would recognize the pattern immediately. ## If you’re in this right now I’ve been doing structural work — information architecture, workflow design, content operations — for a long time. Some of it was called web development. Some of it was called marketing operations. What you’re calling AI implementation today involves the same underlying questions. If your team is figuring out where the agents go and the structure isn’t solid yet, I can help you work through it. Reach out — I’d rather help you get ahead of the problem than pull you out from under it later. _JCody Ventures — video, drone, networks, marketing, and the structure underneath all of it. [Get in touch.](https://jcody.com/contact)_ --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [Frederick Brooks told you this would happen. In 1971.](https://jcody.com/insights/frederick-brooks-mythical-man-month-ai-agents/) appeared first on [JCody Ventures](https://jcody.com). ## I wrote the filing system for AI agents in 2005. I just called it a website. - Link: https://jcody.com/insights/filing-system-for-ai-agents-2005/ - Published: 2026-10-26T11:55:00-04:00 - Author: Jeffrey Cody If you’re building AI agents right now — figuring out what they can see, what they’re allowed to change, how they hand information off to each other — you’re solving a structure problem. I built a version of that structure in 2005. I just called it a website. Not bragging. Making a point: the structure hasn’t changed. The workers have. ## What building websites taught me about information architecture In 2005, I was deep in information architecture — building systems that decided what content went where, who could touch it, what happened when it conflicted. We had a framework: levels of hierarchy, tiers of authority, one source of truth for any given fact. If a piece of content didn’t have a clear home, it didn’t belong yet. That wasn’t a limitation. It was discipline. We called certain pages “orphans” — content that existed but had no clear relationship to the rest of the site. Orphans create confusion. They slow everything down because nobody is sure who owns them or what they’re supposed to do. Sound like any AI workflow you’ve seen lately? ## The same rules apply to agents An AI agent needs to know: what am I reading? What am I allowed to write? Who do I hand off to, and when? What’s the source of truth if two things conflict? Those aren’t AI questions. They’re information architecture questions. The difference is that in 2005, a confused system just showed you a broken page. Today, a confused system makes decisions — sometimes expensive ones — before anyone notices it’s lost. The discipline matters more now, not less. ## What to do before you add another agent Map what you already have. Draw the workflow. Name the source of authority for each type of information in the system. Find your orphans — the inputs or outputs that don’t clearly belong anywhere — and sort those out before you scale. It takes more time up front. It’s dramatically faster once you’re moving, and it avoids the painful rebuild that happens when you’ve already embedded agents into real processes. This is the work I do. If your team is building something and the structure isn’t solid yet, reach out. I’d rather help you get the foundation right than help you untangle it later. _JCody Ventures — video, drone, networks, marketing, and the structure underneath all of it. [Get in touch.](https://jcody.com/contact)_ --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [I wrote the filing system for AI agents in 2005. I just called it a website.](https://jcody.com/insights/filing-system-for-ai-agents-2005/) appeared first on [JCody Ventures](https://jcody.com). ## Drone footage that tells the story: what I shoot, who for, and why it sells - Link: https://jcody.com/insights/drone-footage-that-tells-the-story/ - Published: 2026-10-26T11:40:00-04:00 - Author: Jeffrey Cody Drone footage earns its place in a video when it helps tell the story, not when it shows off the flying. After more than a decade behind a camera, that is the lesson that changed how I shoot and how I plan. ## Who I shoot for * Local landscaping and property businesses that want finished work seen from above. * Churches and community groups documenting a campus or an event. * Riders and families at equestrian shows, with everyone’s consent. * Local public-safety agencies, when a view from above helps someone make a decision. ## What makes aerial footage compelling People do not care how the drone was flown. They care about seeing a familiar place from a view they have never had. The best compliment I can get is not that the flying was great but that someone wants to know how I got the camera through there. When the difficulty disappears, the story is what is left. ## The planning nobody sees A thirty-second flight looks quick. Behind it sit the route, the light, the wind and the wait for the right moment. The gap between a good shot and a great one is often a few centimeters, a change in speed, or ten minutes of patience. I rehearse so the final take looks effortless. ## Where the experience comes from I have held my FAA [Part 107](https://www.faa.gov/uas/commercial_operators/become_a_drone_pilot) remote pilot certificate since December 2021, earned so I could legally shoot enterprise- quality video for a corporation. My first unmanned aircraft work was in 1994, as an intern on a US Army program flying full-size radio-controlled aircraft for parachute testing. The gear has changed. The habit of planning every flight has not. ## Why it sells A client is not buying flying. They are buying the moment their customer, congregation or neighbor sees their place in a new way, and the story that moment helps them tell. That is the footage I try to deliver. If you have a project in mind, [start a conversation](https://jcody.com/contact/), or read how I handle the large video libraries behind it in the [AI projects](https://jcody.com/ai-projects/) section. --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [Drone footage that tells the story: what I shoot, who for, and why it sells](https://jcody.com/insights/drone-footage-that-tells-the-story/) appeared first on [JCody Ventures](https://jcody.com). ## Safe archival compression: reclaim space, keep the raw - Link: https://jcody.com/insights/safe-archival-compression-reclaim-space-keep-the-raw/ - Published: 2026-10-26T11:30:00-04:00 - Author: Jeffrey Cody Archival compression is how I get storage back from a very large video archive without ever putting the originals at risk. With close to 100 TB of footage, some of it impossible to reshoot, I do not take chances. ## The rules I do not break * Originals are never modified. * Compression runs on copies, never on the source. * A quality score decides whether a compressed copy is allowed to stand in for the original. * Nothing is removed until the new copy is verified and I approve it. * Every step leaves a record. ## How the quality gate works I compare each compressed file against its source with an objective quality measure called [VMAF](https://github.com/Netflix/vmaf), and I also check that the new file is meaningfully smaller. If a clip does not clear the bar, the original stays exactly as it was and the file is logged. A shaky drone clip missed the mark once, and the system kept the original, which is the whole point of having a gate. ## Log it, park it, move on One bad file should not stop the line. A file that fails is logged and parked for a human decision, and the run continues. I learned that rule in marketing long before I used it with AI: do not stop the show for one problem, flag it and keep the curtain moving. ## Keep the raw Compressed copies are for working, sharing and everyday use. The raw files stay, for the day an edit needs the full-quality source. Reclaiming space and keeping the raw are not opposites if the process is careful. This sits on top of the [golden record](https://jcody.com/insights/golden-record-indexing-fingerprinting-25-years/), which tells me exactly what each file is before I touch a copy of it. If you have an archive that is out of room, [let us talk](https://jcody.com/contact/). --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [Safe archival compression: reclaim space, keep the raw](https://jcody.com/insights/safe-archival-compression-reclaim-space-keep-the-raw/) appeared first on [JCody Ventures](https://jcody.com). ## The golden record: indexing and fingerprinting 25 years of data, clients and video - Link: https://jcody.com/insights/golden-record-indexing-fingerprinting-25-years/ - Published: 2026-10-26T11:20:00-04:00 - Author: Jeffrey Cody A golden record is the one version of the truth that everything else points back to. I have been building mine from 25 years of clients, projects and video, and it is the most useful thing I have made for finding anything. ## What goes into the golden record Every file gets three things. First, a fingerprint: a cryptographic hash that proves the file has not changed and exposes duplicates even when the names differ. Second, the metadata the camera or software already recorded, such as date, device and duration. Third, tags that describe what is actually in the file, written by local AI and checked by a person. ## Relationships matter as much as files A clip belongs to a shoot, a shoot belongs to a client, and a client belongs to a relationship that may go back years. The record keeps those links, so a search for aerial, church, autumn returns clips instead of folders. Without the relationships you only have a very organized pile. ## One source of authority When two copies disagree, the record says which one wins. Everything else points to it, and nothing quietly drifts apart. I learned that rule building websites in 2005, which is a story I tell in [I wrote the filing system for AI agents in 2005](https://jcody.com/insights/filing-system-for-ai-agents-2005/). ## Why a business should care Most small organizations have this problem at a smaller scale: years of files, and nobody sure which are duplicated and which can never be replaced. You cannot clean up what you cannot see, so I index first and decide second. Originals are never touched while that happens. The golden record is also what makes [safe archival compression](https://jcody.com/insights/safe-archival-compression-reclaim-space-keep-the-raw/) possible. If you are sitting on years of accumulated data and want a plan, [talk to me](https://jcody.com/contact/). --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [The golden record: indexing and fingerprinting 25 years of data, clients and video](https://jcody.com/insights/golden-record-indexing-fingerprinting-25-years/) appeared first on [JCody Ventures](https://jcody.com). ## Infrastructure automation: getting the most out of the hardware you already have - Link: https://jcody.com/insights/infrastructure-automation-get-the-most-from-your-hardware/ - Published: 2026-10-26T11:10:00-04:00 - Author: Jeffrey Cody Infrastructure automation usually gets sold as a shopping trip. In my experience it starts with a measurement. Before you buy a bigger server, a faster network card or another drive, find out which part is actually holding you back. ## Measure before you buy My video archive server is a used Xeon machine with 64 GB of ECC memory and a 10- gigabit network connection. On paper it is modest. In practice the thing slowing a job down was rarely the part I would have guessed. Sometimes it was the disks, sometimes the network path, sometimes one slow step that everything else waited on. A few minutes with the right monitoring saved me from buying the wrong upgrade more than once. ## Fix the bottleneck in the right order Fix one limit, then measure again, because the next limit is now the new problem. Change one thing at a time so you know what actually helped. It is slower for about a day and much faster for the next year. ## Automate the repeatable parts Hashing files, checking disk health, running backups, tagging footage and compressing copies are the same steps every time, so they belong to scripts and scheduled jobs. A script does not get bored or skip a step on a Friday afternoon. My rule is simple: scripts do the work, and a person approves anything that changes or removes data. ## Right-size the hardware Matching components to the job beats chasing the biggest number. For video work that can mean a graphics card with enough memory and the right encoding support for the footage you shoot, not simply the newest card on the shelf. I am choosing mine by what the workload needs, and I would suggest you do the same. If a team you know is running on hardware that merely feels slow, the cheapest first step is to find out why. The same approach drives how I [manage networks and deploy change](https://jcody.com/insights/managing-networks-deploying-change-faster-safer/), and I am happy to look at yours: [start a conversation](https://jcody.com/contact/). --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [Infrastructure automation: getting the most out of the hardware you already have](https://jcody.com/insights/infrastructure-automation-get-the-most-from-your-hardware/) appeared first on [JCody Ventures](https://jcody.com). ## Managing networks and deploying change faster and safer - Link: https://jcody.com/insights/managing-networks-deploying-change-faster-safer/ - Published: 2026-10-26T11:00:00-04:00 - Author: Jeffrey Cody Managing networks well comes down to one habit: never make a change you cannot take back. Everything else I do follows from that. ## Know what you have before you touch it Every network I inherit starts the same way. Someone built it years ago, it mostly works, and nobody is quite sure why. Before I change anything I write it down: the gateway, every switch and access point, which port feeds which wall jack, and what lives on which segment. It is the dullest part of the job and it prevents the worst phone calls. ## Segment so one mistake stays small Cameras, door access, guest Wi-Fi, office computers and a livestream encoder should not share one flat network. I split them apart so a noisy device cannot take down the front desk. On a church campus I rebuilt with Ubiquiti UniFi gear, that separation is what lets the cameras, the door locks and the people in the office each do their jobs without stepping on one another. ## Script the checks, not the fixes Before and after every change I run the same short list: can the gateway see each switch, are the cameras recording, does the VPN come up, is DNS answering. A script does it in seconds and writes the result to a log, which beats my memory at the end of a long day. Automation can watch the network all day. A person decides what changes. ## Make every change reversible Back up the configuration. Change one thing. Test it. If it is wrong, put it back. I write the rollback note before I make the change, not after it goes sideways. The same rule applies at home, where my own gateway and VPN let me reach the archive server from the road. One careless step there could lock me out of my own building, so I make those changes one at a time too. ## Faster comes from safer It sounds backward, but the more reversible and checked each change is, the faster I can move. You stop dreading changes, so you make small ones often instead of one big frightening one every few years. If you are running a network nobody wants to touch, that is usually the right place to start. The same discipline protects my video archive, which you can read about in [safe archival compression](https://jcody.com/insights/safe-archival-compression-reclaim-space-keep-the-raw/). If you would like a second set of eyes on your own setup, [get in touch](https://jcody.com/contact/). --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [Managing networks and deploying change faster and safer](https://jcody.com/insights/managing-networks-deploying-change-faster-safer/) appeared first on [JCody Ventures](https://jcody.com). ## Free Range 2.0: the little girl grew up - Link: https://jcody.com/insights/free-range-2-0/ - Published: 2026-10-26T10:00:00-04:00 - Author: Jeffrey Cody A while back I wrote about [free-range kids and the Cody way](https://jcody.com/insights/free-range-kids-the-right-way/): let them try things, let them find their own limits, stay close enough to catch the big falls. I owe you the update, because that little girl in the story is not little anymore. ![Piper clearing a jump on her chestnut hunter at the Saratoga horse show](https:// jcody.com/wp-content/uploads/2026/10/jcody_piper-saratoga-hunter-jumper_2026.jpg) Piper over the fence at Saratoga. Piper rides hunter/jumpers. If you have never watched it, the sport is a lot of patience, a lot of repetition, and about eight seconds of something that looks effortless. It looks easy because somebody did the boring part a thousand times first. That is free range, grown up. Nobody can hand a kid a sixteen-hundred-pound partner and a fence and promise it will be safe. What you can do is build the reps, the judgment and the habit of getting back up, and then step back. We drive. We hold the gate. We pay attention. She rides. ## The lessons did not change * **Limits come from experience.** The best riders are not the fearless ones. They are the ones who have learned exactly where their edge is. * **Small mistakes are cheap.** A rail down in a schooling round teaches more than a perfect round ever will. * **Help when it is asked for, or clearly needed.** Good coaches know the difference. So do good parents, most days. ![Piper standing with her chestnut horse under blossoming pear trees, the horse resting its head against her shoulder](https://jcody.com/wp-content/uploads/2026/ 10/jcody_piper-with-her-horse-spring-blossoms_2026.jpg) Quiet day, same partnership. The work happens early and mostly out of sight. The pictures show the part that is easy to love. What they do not show is the 3 a.m. alarm, the cold barn and the long drive home. JoAnn said it better than I can, from the rail at a show this summer: > This is what it’s all about. > Not the 3 a.m. wake-ups. Not the cleaning of boots and tack. Not the endless hours > driving to the barn. > This. > Waiting to hear how your heart and soul did with their best friend after they > rode their hearts out. Watching them chase their dreams, give everything they > have, and come back knowing they gave it their all. > This is why we do it. > JoAnn Cody ![JoAnn watching a horse show ring from the rail as riders and horses warm up on the footing](https://jcody.com/wp-content/uploads/2026/10/jcody_joann-watching-the- ring-horse-show_2026.jpg) After all, it’s why we are here. JoAnn is the horse person in this house, and a lot of what we know about [patience, preparation and showing up on time](https://jcody.com/capabilities/equestrian/) comes from there. It is also why the rest of JCody Ventures is built the way it is: do the unseen work well, so the day itself looks easy. --- _Written with AI assistance. Every idea, opinion and experience here is mine. I just thrive on strategy and execution more than eloquent wordsmithing._ The post [Free Range 2.0: the little girl grew up](https://jcody.com/insights/free-range-2-0/) appeared first on [JCody Ventures](https://jcody.com). ## Selling styles: art or science (a love story) - Link: https://jcody.com/insights/selling-styles-art-or-science-a-love-story/ - Published: 2026-07-21T09:00:00-04:00 - Author: Jeffrey Cody I spent years creating leads for other people’s sales teams, and one thing bugged me every time. Two reps get leads of the same quality. One closes 18 percent. The other closes 8. That gap is not luck, and it is not all talent. Most of the time, the best rep is doing something the others are not, and nobody has written it down. ## Art or science? Yes. Selling is both. But the more of the art you can capture, the more of it you can teach. Writing down what your best person does turns private skill into a process anyone on the team can run and everyone can measure. “Process” is not a dirty word. Every sales team has one, written or not. Written processes just get better, because you can see them. ## Who does what * **Executives** fund the tools that let managers watch, measure and coach. * **Managers** build the best-practice process from what actually wins. * **Reps** follow it, so results mean something, then help improve it. ## Look for patterns Track where deals are won and lost. Find what the top rep does differently. Collect it in custom CRM fields, because you cannot decide on data you never captured. When it works, weaker reps start closing like stronger ones, managers hit their numbers, and the lead-generation budget stops being a leak. ## The 2027 mentality The best reps are now being copied by software. Call recordings and CRM notes can be read at scale, and the patterns that used to take a manager a quarter to spot show up in a week. I would use that in two ways: coach with the evidence instead of the anecdote, and let reps use AI for prep and follow-up so they spend more of the day on the part machines cannot do, which is being a person across the table. The post [Selling styles: art or science (a love story)](https://jcody.com/insights/selling-styles-art-or-science-a-love-story/) appeared first on [JCody Ventures](https://jcody.com). ## Have you unoptimized your marketing campaigns yet? - Link: https://jcody.com/insights/have-you-unoptimized-your-marketing-campaigns-yet/ - Published: 2026-06-16T09:00:00-04:00 - Author: Jeffrey Cody Marketers love to optimize. Test the subject line, narrow the segment, trim the audience, swap the creative. All good habits. But if you keep tightening for long enough, you can wring out the opportunities that were quietly working. Call it over-optimization. The fix is to _unoptimize_, on purpose and in small doses. ## A simple exercise Pretend you are launching your current campaigns today. Ask yourself: * Has the product changed since we set this up? * Has the market, or the words customers use for what we sell? * Has the team, the sales feedback or the creative style changed? * Is the audience at a different stage of readiness than it was? A decision you made 6 to 18 months ago may have been right then and wrong now. Somewhere in there is an audience or an offer you cut and never revisited. ## Do not tear everything down Pick a few past cuts and reopen them one at a time, with a small budget and a clear measure. That finds the missed opportunities without putting your whole return at risk. Then bring your managers and lead-generation people into the review. They will remember what you forgot. ## The 2027 mentality Every campaign should carry its own memory: what we cut, why, and what we expected to see. AI is good at keeping that log and good at resurfacing it on a schedule (“ you excluded this segment 11 months ago, here is what changed”). Let it keep the notes and run the cheap re-tests. Keep your judgment for deciding what is worth trying again. The post [Have you unoptimized your marketing campaigns yet?](https://jcody.com/insights/have-you-unoptimized-your-marketing-campaigns-yet/) appeared first on [JCody Ventures](https://jcody.com).