I've spent 15 years in supply chain management. I've managed procurement for a $400M manufacturer, led logistics optimization for a 3PL, and overseen warehouse operations across 12 facilities in three countries. I've seen every spreadsheet hack, every workaround, every system that was supposed to "transform" our operations and didn't.
I'd thought about starting a company before. Vaguely. The way you think about learning guitar or running a marathon — aspirational, never urgent. I had a good salary, a 401k, and a title that took a decade to earn. The startup world felt like it belonged to 24-year-olds with computer science degrees and a willingness to eat ramen.
Then a colleague sent me a link to Vantage. "It's this AI thing that interviews you about your career and finds startup ideas," she said. "Took me 15 minutes."
I figured it would spit out generic suggestions. "Start a consulting firm." "Build an app." The kind of advice you get from a career coach who's never worked in your industry.
I was wrong.
The Interview
The experience wasn't what I expected. I'd assumed it would be a form — checkboxes, dropdowns, the usual. Instead, it was a conversation. The AI asked me about my day-to-day work, and then dug deeper.
"What's the most frustrating part of your current workflow?"
I started talking about vendor compliance documentation. How every new supplier requires 15–30 documents — certificates of insurance, quality certifications, sustainability attestations, tax forms — and how we track all of it in a shared drive with a naming convention that nobody follows. I told it about the time we nearly lost a $2M contract because a supplier's insurance certificate had expired and nobody caught it until the client audit.
"How do you currently handle that process?"
I described the spreadsheet. Fourteen tabs. Color-coded expiration dates that our procurement coordinator updates manually every Monday morning. The monthly email chain where we chase suppliers for renewed documents. The annual panic when audit season arrives and we're scrambling to locate certificates we swore we had.
"How many hours per week does your team spend on this?"
I'd never actually calculated it. After thinking about it: roughly 20 hours per week across three people. That's over 1,000 hours per year — $75,000 in fully loaded labor cost — just tracking documents that expire and need renewal.
The AI kept going. It asked about procurement bottlenecks, about what information I wish I had that I don't, about which vendor meetings I dread and why. It asked about my relationships with suppliers — how I built them, what made the good ones work, why the bad ones failed.
Twenty minutes in, I'd said more about my industry frustrations than I'd said in 15 years of performance reviews.
The Five Ideas
When the analysis came back, it wasn't a generic list. Each idea was tied directly to something I'd described — a pain point I'd been living with so long I'd stopped seeing it as a problem.
Idea 1: Automated Vendor Compliance Management
The connection: My spreadsheet rant.
The AI identified that my 14-tab spreadsheet wasn't a personal quirk — it was an industry pattern. Across manufacturing, food production, construction, and healthcare, procurement teams manage vendor compliance documentation with manual tracking systems. The market for vendor compliance management is estimated at $2.4 billion, growing at 18% annually as supply chain due diligence requirements intensify.
The specific angle it suggested: an AI-powered platform that ingests vendor documents (via email forwarding or supplier portal), extracts key data (expiration dates, coverage limits, certification scopes), flags gaps against company-specific requirements, and auto-requests renewals 60 days before expiration.
I thought about it. I know 40+ procurement managers who have the same spreadsheet. I could describe their exact pain because it's my pain. I know what the existing solutions get wrong — they're either enterprise-grade (SAP Ariba, Coupa) at $100K+/year or they're generic document management that doesn't understand compliance requirements.
My honest reaction: This was the most obvious idea I'd never had. I'd spent 15 years managing this spreadsheet and never once thought, "I should build the thing that replaces it."
Idea 2: Supplier Relationship Intelligence
The connection: My answer about which vendor meetings I dread and why.
I'd described how the best supplier relationships are built on consistent communication, fair treatment during negotiations, and mutual transparency. But I'd also described how we have no system for tracking relationship health — it's all institutional knowledge in people's heads. When our best procurement manager left last year, we lost all context on 200+ supplier relationships overnight.
The idea: a CRM built specifically for supplier relationships (not customer relationships — those are entirely different dynamics). Track communication frequency, negotiation history, quality performance, delivery reliability, and relationship risk scores. Alert when a critical supplier relationship shows signs of deterioration — delayed responses, declining quality metrics, contract pushback.
Market context: The supplier relationship management market is $3.1 billion, but most solutions are modules within larger procurement suites, not standalone tools built for the actual relationship dynamics.
Idea 3: Freight Cost Anomaly Detection
The connection: My description of invoice auditing.
I'd mentioned that we spot-check freight invoices because auditing all of them is impractical. The AI latched onto this: if you can only review 15% of invoices, you're statistically likely to miss $200K–$500K in annual overcharges for a company shipping $30M+ in freight.
The idea: an AI layer that sits between carrier invoices and payment processing, automatically flagging anomalies — charges that don't match contracted rates, duplicate billings, incorrect accessorial charges, fuel surcharge miscalculations. Not a full freight audit platform (those exist and are expensive), but a lightweight anomaly detection layer that integrates with existing TMS and ERP systems.
The insight I wouldn't have reached alone: I knew we overpaid on freight. Everyone in logistics knows this. But I'd framed it as "an annoying cost of doing business," not "a $2.6 billion market opportunity in freight audit automation." The reframing from personal frustration to market opportunity was the shift I needed.
Idea 4: Warehouse Labor Demand Forecasting
The connection: My description of seasonal staffing chaos.
Every Q4, my warehouses need 40–60% more labor. Every Q4, we scramble. I'd described the process: reviewing last year's volumes (which are never the same as this year's), calling temp agencies, hoping people show up, training new workers in 2 days instead of the 2 weeks they need, watching productivity drop 30% while error rates triple.
The idea: an ML model trained on historical order data, promotional calendars, weather patterns, and economic indicators to forecast warehouse labor demand 6–8 weeks out. Integrated with temp staffing platforms to auto-request workers when demand is predicted to spike. Include a training module that onboards temp workers before they arrive.
Why it resonated: I've spent a decade wishing I could see the future of my labor needs. The data to do it exists — order patterns, promotional schedules, seasonal trends — but nobody has assembled it into a predictive tool for mid-market warehouses. Enterprise WMS systems have basic forecasting, but the 180,000+ warehouses under 200,000 sq ft have nothing.
Idea 5: Supply Chain Risk Monitoring for Mid-Market Companies
The connection: My description of COVID-era supply chain disruptions.
I'd talked about how we were blindsided when a key supplier's factory in Guangdong shut down during COVID lockdowns. We had no visibility into our tier-2 and tier-3 suppliers. It took us 8 weeks to find alternative sourcing — 8 weeks of production delays, expedited shipping costs, and customer penalties.
The idea: a monitoring platform that maps multi-tier supply chains and continuously scans for risk signals — factory shutdowns, port congestion, regulatory changes, financial distress indicators, extreme weather events, and geopolitical tensions — then alerts procurement teams with severity scores and recommended actions.
Market context: Enterprise supply chain risk tools (Resilinc, Everstream) cost $200K–$1M annually. Mid-market manufacturers ($50M–$500M revenue) need the same intelligence at 10% of the cost. The supply chain risk analytics market is projected at $4.3 billion by 2027, with the mid-market segment almost entirely unserved.
What I Realized
Five ideas. All of them mapped directly to problems I deal with every single day. Problems I'd been solving with spreadsheets, phone calls, and institutional knowledge for 15 years.
I hadn't missed these opportunities because I was unimaginative. I'd missed them because when you're inside the system, the system's failures feel normal. The spreadsheet isn't a problem — it's just how things work. The annual staffing scramble isn't a market opportunity — it's just Q4. The freight overcharges aren't a product — they're a cost of doing business.
The AI interview did something I couldn't do for myself: it asked me the right questions, listened to my answers without the bias of familiarity, and reframed my daily frustrations as market opportunities with quantifiable demand.
I'm now three months into validating Idea 1 — vendor compliance management. I've talked to 60+ procurement managers. Every single one has a version of my spreadsheet. Forty-three of them said they'd pay for a solution. Twelve are on a waiting list for the beta.
I haven't quit my job yet. I'm building on evenings and weekends, using the part-time founder approach. My domain expertise means I can build the right thing the first time — I don't need to "discover" the problem through 50 pivot cycles. I already know what the product needs to do because I've needed it for a decade.
My 15-year career wasn't a detour from entrepreneurship. It was preparation for it. I just needed something to show me that.
Discover startup ideas hidden in your career experience — take the free AI interview →