David Okafor met his co-founder at a Lagos tech meetup in 2022. They were both building in public—tweeting about AI projects, sharing GitHub repos, comparing notes on API costs. The chemistry was immediate.
By January 2024, they’d incorporated. By March, they’d raised $150,000 pre-seed. By June, they had 12 beta customers for their AI customer support platform.
By September, Okafor was building alone. His co-founder—let’s call him Tunde—had simply stopped responding to messages. Not a dramatic exit. Not a disagreement. Just silence. Slack unread. WhatsApp blue ticks, no reply. Email bouncing.
The company survived. Okafor didn’t, not initially. This is the story of building through the specific trauma of abandonment—and the systems he built to ensure he’d never be this vulnerable again.
The Partnership That Wasn’t
Okafor and Tunde split roles cleanly: Tunde handled sales and investor relations, Okafor built the product. Classic division. Classic vulnerability.
The warning signs Okafor ignored:
- Tunde’s commits to the GitHub repo had slowed to 1–2 per week by May
- Investor updates were increasingly written by Okafor, signed by Tunde
- Tunde started “working from home” more frequently
- The shared calendar showed increasing “personal appointments”
Okafor rationalized each. Tunde was networking. Tunde was fundraising. Tunde was managing his energy.
The truth was simpler. Tunde had discovered the company was harder than expected. The customers were demanding. The technology was brittle. The competition from Zendesk and Intercom was crushing. He was slowly withdrawing, preparing his exit without the confrontation of actually quitting.
The Day of Silence
September 14, 2024. Okafor sent a routine message about a customer escalation. No response in 4 hours. Unusual but not unprecedented. He sent a follow-up. Nothing. He called. Voicemail. He messaged Tunde’s wife, a friend. “He’s fine, just stressed, needs space.”
The space lasted 3 weeks. Then an email from Tunde’s lawyer: voluntary departure from the company, equity forfeiture per the vesting agreement, no further communication requested.
Okafor later learned Tunde had taken a product management job at a bank. Stable salary. No equity. No customers. No 3 AM outages.
The betrayal wasn’t the departure. It was the method. No conversation. No transition. No acknowledgment of the 18 months they’d spent building together. Okafor had been erased from Tunde’s narrative, and the erasure itself was the wound.
The Immediate Collapse
Okafor spent October 2024 in paralysis. He’d never spoken to investors alone. Never done a sales demo. Never written a marketing email. The division of labor that had enabled scale had also prevented him from learning half the business.
The company metrics during his absence:
| Metric | September | October | November |
|---|---|---|---|
| Active customers | 12 | 9 | 6 |
| MRR | $4,800 | $3,600 | $2,400 |
| Support response time | 2 hours | 14 hours | 3 days |
| Churn rate | 5% | 25% | 40% |
The product worked. The customers were leaving because the founder had disappeared into grief.
Okafor’s mother found him in his apartment on November 3rd, after 4 days of unanswered calls. He’d been sleeping 14 hours daily, eating delivery food, refreshing Tunde’s social media to see if he’d posted about the new job. He hadn’t. The erasure was complete.
The Rebuild, With Scars
Therapy helped. Specifically, a Lagos-based psychologist who specialized in “high-functioning trauma”—the kind that doesn’t stop you from working but poisons the work’s meaning.
The first insight: Okafor had merged identity with partnership. “We were going to build something” had become “I am only real through this relationship.” The ghosting triggered an existential collapse, not just a business one.
The second insight: his technical skills were real, validated by 12 customers who’d chosen his product over alternatives. The business was salvageable. He was salvageable.
Okafor spent December 2024 doing three things:
1. Customer resurrection
Personal calls to every churned account. No sales pitch. Just: “I failed you. Here’s what happened. Here’s how I’m preventing it from happening again.” Three customers returned. Two referred others.
2. Structural hardening
- All critical passwords in 1Password, with board member access
- Weekly investor updates, written by him, no matter what
- No single point of failure in any process
- Vesting agreements reviewed with new counsel
3. Solo operational training
He spent 2 weeks shadowing a sales coach, 1 week with a marketing freelancer, 3 weeks doing every support ticket himself. The skills Tunde had handled were now his, imperfectly but present.
The New Company, Same Product
By March 2025, Okafor had renamed the company (new identity, clean break) and restructured the offer. The product—AI customer support—was unchanged. The business around it was unrecognizable.
The new structure:
| Function | Previous | Current |
|---|---|---|
| Leadership | Co-CEO model | Solo founder + advisory board |
| Sales | Co-founder driven | Automated demo + founder closes |
| Support | Co-founder handled | AI-first, human escalation |
| Marketing | None | Content-driven, Okafor’s voice |
The advisory board included his lead investor, a former founder who’d survived a similar co-founder departure, and a psychologist who understood founder mental health. They met monthly. They asked about his sleep, his relationships, his sense of meaning.
The Specific Systems That Prevent Recurrence
Okafor now teaches these to other solo founders:
The “Bus Factor” Audit
Every quarter: if I disappear for 30 days, what breaks? Fix the top 3.
The “No Heroes” Documentation
Every process must be runnable by someone else within 2 hours of reading. No “Tunde knows how to do this.”
The “Employment Continuity” Agreement
Co-founders sign: if you leave, you provide 30 days transition or forfeit accelerated vesting. Legal enforcement of moral obligation.
The “Solo Founder” Support Structure
Weekly therapy, monthly peer group, quarterly “existential check-in” with advisor. Building alone doesn’t mean processing alone.
The Current Reality
As of April 2026, Okafor’s company—now called Resolvr—has 34 customers and $18,000 MRR. He’s raised a small seed extension ($400,000) on terms he understands completely. He’s hiring his first employee in May: a customer success lead, not a technical co-founder.
He still checks Tunde’s LinkedIn occasionally. Tunde is a senior product manager at the bank, no mention of the startup years. The erasure remains complete.
Okafor’s final insight: “I used to think co-founders were necessary for the technical stuff. Now I think they’re necessary for the emotional stuff—but only if they stay. I’d rather build systems for the technical parts and communities for the emotional parts. They’re more reliable than any single person.”
Kivora documents founder relationships and the psychology of building alone. Join the newsletter — Subscribe here.
