In this article
On October 6, exactly three years after incorporating DOSS, we hosted DOSSCON1 - our first community summit in San Francisco. We brought operators, executives, and technology into a unified conversation about the hard problems and real-world tactics of scaling businesses today. The conversations revealed something bigger than tactical advice: we’re living through a fundamental platform shift in how businesses operate, from ERP (Enterprise Resource Planning) to ARP (Adaptive Resource Platform).
As a demonstration of this shift onstage, DOSS co-founder Wiley Jones unveiled dossbot—an AI copilot built directly into DOSS ARP that can query data, identify errors, generate workflows, and automate operations through chat prompts. It’s changing how operations teams work: from manually configuring systems to conversationally directing them.
To understand why dossbot matters, you need to understand the problem it solves.
The Problem That Won’t Go Away
“ERP is dead. Long live ARP.”
The story of enterprise software has evolved alongside the story of computing. From 1950s mainframes crunching numbers to SAP’s early MRP software in the 70s to NetSuite’s cloud ERP in the 90s, each platform shift unlocked new markets by slashing costs and expanding access. History is repeating itself again, as AI unleashes the full agility of the value chain.
Because here’s the uncomfortable truth: despite hundreds of billions of dollars spent annually on ERP software, implementation services, and consulting, the results are underwhelming. 50% of executives estimate their ERP programs delivered negative ROI. 75% of implementations fail. 80% go over budget and deliver late.
It's not you. It’s actually just really hard. Why? Because modeling a value chain requires solving three nearly impossible challenges simultaneously: high dimensionality in the last mile (every business is unique), the stakes are incredibly high (systems can’t go down), and businesses change faster than systems can adapt.
Former NetSuite leaders Eileen Tobias and Ravi Ramakrishnan validated this from the trenches. “After 30 years of doing this, implementation still takes six months to a couple years,” Ravi noted. Success requires more than software; it requires “a degree of courage” from leadership.
What the Pioneers Got Right
Fifty years of failed ERP implementations taught the industry some hard lessons. The most successful ERPs recognized something fundamental: master data and transactional data follow the same patterns, just on different dimensions. SKUs and suppliers get referenced in POs and sales orders, which flow into inventory reports. Accounts and departments go in journal entries, which flow into balance sheets and P&L statements.
The engineers who built SAP, Oracle, and NetSuite created dimensional rules engines that bound master and transactional data together through an immutable ledger: the source of truth that ensures your physical workflows and financial workflows never drift out of sync.
This architecture worked. It still works. But it was built on a nearsighted assumption: businesses would mold their strategy to fit what the applications could do, and not need to adapt over time. Traditional ERPs also left the hard work of data modeling to customers, treating catalogs as fixed artifacts rather than living, evolving entities. “Use the system out of the box as much as you can,” is the mantra of every implementation consultant.
The NetSuite panel identified three customer archetypes: “do it for me” (hands-off, often struggles), “I'll do it myself” (DIY, frequently fails), and “let’s do it together” (collaborative, usually succeeds). As Eileen put it: “If you’re not willing to put the work in, you’ll probably dislike your new system just as much as you disliked the old one.”
The problem? Most companies can’t afford 6-18 month implementations anymore. The global value chain evolves too fast and the old model breaks.
When Reality Hits: A Coffee Story
Anthony Fassio, Chief Retail & Operations Officer at Verve Coffee Roasters, lived this tension. After 18 years of organic growth building a beloved brand, Verve’s operational infrastructure was “spreadsheets on spreadsheets on spreadsheets.”
Then came the Whole Foods moment. Sales that had been steady at 35 units per week suddenly exploded to 510—a perfect storm of Verve’s internal promo, Whole Food’s promo, and Amazon Prime Day all at once. “We ran out of stock,” Anthony admitted. “It was super stressful.”
The problem wasn’t just production—it was distribution, forecasting, and data accessibility compounding together. “We have all the data, but we don’t have it in a way where it can inform our managers,” Anthony explained. “Without real-time information, it comes down to either intuition or you run around like a chicken with its head cut off.”
His focus now: brutal honesty about margins and overindexing on technology rather than just hiring more people. His biggest regret? Not implementing rock-solid financial accounting earlier. “The product will be there if your founder is passionate. Operations will figure itself out. But you turn around one day and realize there’s no financial accounting. When you put fuel on the fire to grow, that’s where it falls apart.”
The New Rules of Winning
Nate Moyer, Co-Head of Disruptive Commerce & Internet at J.P. Morgan, opened his session with a reality check: the top 10% of income earners account for 50% of consumer spend, while the bottom 60% account for just 20%. The investment landscape has fundamentally shifted—companies now need visibility to profitability within 12-18 months (down from 24), strategic investors want $100-200M+ in revenue (up from $25-50M), and many companies are down 70-95% from peak valuations.
“Investors have gone back to a back-to-basics model,” Nate explained. “You must be profitable.”
The companies winning today move fast, understand true margins, and make data-driven decisions in real time. Nate highlighted Once Upon a Farm, which went public with 40% of sales from digital—4x the industry average. “You’re getting all that customer data. You’re smarter about them. You can target the next product to them,” he noted.
On AI, Nate identified three game changers: AI-driven content for targeting different customer segments, predictive analytics to solve inventory nightmares (hero SKUs constantly out of stock, non-hero SKUs gathering dust and killing margins), and hyper-personalization at scale. “If everybody else starts getting good at this, it’s going to be a problem,” he warned.
But AI can’t fix broken data infrastructure. You can’t hyper-personalize when your systems don't talk to each other. You can’t predict demand when your forecasting lives in spreadsheets. You can’t be profitable in 12-18 months when implementing new systems takes just as long.
The brands that will command premium valuations aren’t just the ones with great products—they’re the ones with fast, adaptive operational infrastructure.
The Platform Shift: Built for AI from the Ground Up
This is why the shift from ERP to ARP matters. DOSS was built on a different philosophical axiom: design software around what your business actually does, instead of building your operations around what your software can do.
When we founded DOSS three years ago, we knew we had to get three things right:
- Composable – to address the last mile complexity that makes each business unique
- Durable – so you’re not “vibe coding” your accounts payable
- Accelerated – because 6-18 month implementations aren’t fast enough, and it shouldn’t take months to make minor changes to the system
Composable? Yes. Durable? Check. Accelerated? Not enough... until now. DOSS already deploys in weeks, and lets you reconfigure tables and workflows in minutes with a few clicks. But here’s the thing: clicks are still manual. Making the system fast isn’t enough; we need to make the people using it faster too.
That’s where dossbot comes in. This isn’t AI bolted onto rigid, hardcoded ERP data structures; dossbot only works because DOSS was designed for AI from the ground up. By moving schema control to the application layer, creating fully composable tables and workflows, and unifying master data through rigorous onboarding that eliminates duplicates and codifies data governance, DOSS creates a clean, organized, adaptable data architecture that LLMs can actually work with effectively.
Wiley demonstrated the magic onstage. He queried hundreds of thousands of records, identified data errors, generated new table views, summarized trends and insights, and created a diagram of an entire data model—all through dossbot chat prompts.
“We’re only just now giving it spoons,” he joked, referencing the Shawshank Redemption prison escape. “When we give it more tooling, you’re not going to get ASCII diagrams—you’re going to get a full visualization you can roam around in like Matthew McConaughey at the end of Interstellar.”
Dossbot is currently in closed beta. The roadmap ahead includes workflows that compile to code, version-controlled configurations (goodbye multi-year upgrade projects), and dossbot automating everything from 3-way matches on payables to EDI trading partner setup to forecast generation.
What This Means for Operators
As we wrapped up DOSSCON1, several themes emerged across all the conversations:
- Data exists; accessibility is the problem. Companies have data, but they can’t effectively access it, trust it, or act on it fast enough.
- Margins matter more than ever. In a world demanding profitability within 12-18 months, knowing unit economics by SKU, channel, and customer is non-negotiable.
- AI is coming, but trust takes time. Start small, build confidence, scale. Winners will adopt AI early with the right infrastructure, designed intentionally for LLM.
- Codify before you scale. If you haven't documented and standardized operations, you can’t scale reliably.
- Implementation is the product. The system you buy is only as good as the transformation you commit to, and your ability to choose systems that evolve with you.
The discussions at DOSSCON1 weren’t just about software—they were about the future of how businesses operate. We’re at an inflection point. The winners won’t be the ones with the biggest implementation budgets or the longest timelines. They’ll be the ones who recognize that the platform shift is already here, and the time to adapt is now.
DOSSCON1 was just the beginning. Stay tuned for on-demand session videos dropping soon and follow us for more insights on building resilient, adaptive operations.
