Skip to main content

Call Summary: DeepTrack Strategic Partnership and Co-founder Exploration

Overview​

Gary Sheng, facilitated by Brenda Gentry, engaged in a strategic discussion with Bryan Koyundi, founder of DeepTrack, to explore a potential co-founder or strategic advisor relationship. The conversation revealed DeepTrack's pivot from using third-party APIs to building proprietary AI models for deepfake detection, Bryan's frustration with the African market, and the opportunity for Gary to help penetrate the US market before Bryan's Silicon Valley residency in December 2025.

Key Topics Discussed​

Opening Spiritual Alignment​

Gary's Framework:

"When I think about early stage initiatives that I could get involved in, I feel like it has to be the case where it feels like that's what's happening, where everyone involved is being in obedience, and that's why they're getting involved."

Brenda's Confirmation:

  • Revealed that Bryan had said "God is the co-founder" the day before
  • Gary independently stated he wanted "God as the co-founder"
  • Brenda: "The funny thing is, he said yesterday, God is the co-founder. And you're just saying you want God as the co-founder."

Bryan's Affirmation:

"Yeah, it's true. It's actually true. Very, very true."

Technical Evolution and Competitive Advantage​

Pivot from Grok to Proprietary Models:

  • Initial Approach: "Grok was only giving us about 56% accuracy"
  • Reality Defender Experiment: "Reality Defender was good, but it was very, very expensive. We could not keep up"
  • Current Solution: Building own internal models with 92% accuracy

Intel FakeCatcher Breakthrough:

"We had to actualize Intel FakeCatcher white paper... I saw it on YouTube, reached out to the project manager, and she said it's not going to be available until 2030. The reason is no funding for it at the moment."

  • Intel gave permission to actualize their paper
  • Brought in Dr. Tawil from Morocco (two PhDs in computer vision and AI)
  • Achieved 60% actualization within days

Model Architecture Strategy:

  • 26 Total Models Built:
    • 16 models for insurance alone
    • 8 models for finance
    • 2 models for deepfake detection
  • Philosophy: "Each and every other use case, we have to keep building newer models for that"
  • Cost Advantage: Reality Defender charges $0.40/scan (50 Kenya shillings), DeepTrack built internal IP to avoid this

Market Challenges and Strategic Pivot​

African Market Frustrations:

"It's not an interesting market. It's so hard to build in Africa, and the returns are critically low. It takes so much time to convert."

Specific Examples:

  • Zepri deal: 5 months to close
  • Maurice's experience: Fastest deal was 3 months (CEO was friend), others took 2 years
  • Corruption issues: "You have to grease so many people along the pipeline"
  • Tax burden: "If you would invest today for maybe $10,000, by the time it gets to you, it's probably 600 bucks"

Cultural Buying Patterns:

"Down in Africa, we tend not to buy African products. It's very, very hard to sell an African product here, unless if you're willing to grease and you're willing to bribe... But if they have this idea that, okay, this is probably a US origin product or organization, they'll put their hands on it, they'll immediately buy."

Competitive Intelligence and Positioning​

Reality Defender Validation:

Bill from Reality Defender: "He was telling me that we are only eight months or ten months old, and we seem to have been in existence for two years. In fact, he was even surprised that we are eight months old."

Truepic Parity:

"When I reached out to Truepic, we were solving the same use cases, the same models that we have is what also Truepic has... I didn't even see the need for us now to even do maybe a partnership or borrow their API, because we're literally just doing the same thing as Truepic."

Market Opportunity:

  • Reality Defender: New York-based, expanding to San Francisco and D.C.
  • Truepic: Series B/D funded, focused on insurance
  • DeepTrack: 8 months old, already matching capabilities at fraction of cost

Gary's Value Proposition and Background​

Core Mission Statement:

"I think God created me to make niche, otherwise boring technical things into like exciting ecosystems."

Relevant Experience Highlights:

  • Google Cloud software engineer (2015-2019)
  • Grew Facebook page on police brutality to 800,000 followers, 1 billion impressions
  • Booked Kygo's first US show, built tropical house music movement (110,000 SoundCloud subscribers)
  • Gitcoin Passport growth (100,000 to 1 million users)
  • Naval Ravikant's AirChat early adopter and evangelist
  • Edge City Denver lead organizer (3,000 attendees)
  • Regular White House invitations
  • Upcoming Rome AI conference at Catholic Church

Direct Offer:

"I'm really good at sensing into talent and making niche things... I'm definitely interested in seriously exploring working together."

Business Development Opportunities​

Dating App Integration:

  • Bryan DMed Justin (Tinder co-founder) about deepfakes and catfish profiles
  • "If any deepfake detection company would tap into such a market, that's huge revenue already"
  • Gary's connection: Naval's co-founder from AirChat was Tinder's first CPO

Enterprise Categories Needing Solutions:

  • Social media platforms
  • Dating apps
  • Government agencies
  • Media houses
  • Financial services
  • Celebrity protection

Catholic Church Integration:

Gary: "Imagine DeepTrack being integrated in different ways in Catholic infrastructure around the world."

Silicon Valley Residency Program​

Program Details:

  • Starting December 2025
  • 6-month duration (120 days mentioned)
  • Top 5 selection = free residency with meals and accommodation
  • $10,000 bond for 6 months
  • Weekly pitch competitions
  • Investor fire chats aligned with cybersecurity and AI verticals

Strategic Importance:

Brenda (earlier context): "I don't think the conversation will be the same when he's in the States. That's why I'm trying to get us to secure something for you before."

Future Vision and Market Timing​

Industry Outlook:

"We're still in the very early days of AI-generated content for enterprise. But we still have another two years marks before it completely blows up."

Geographic Market Analysis:

  • US: "Most of the market will still remain in the US"
  • China: "That's also where most of engineering takes place"
  • Europe: "Will not have so much activity... most of it is just regulations and policy"
  • Africa: "Maybe five years later" for market explosion

Agent Integration Vision:

"I honestly believe that in the next 10 years, we're going to see very large companies with a very, very small payroll because... the agent is running at least 50% of the organization."

  • Plans to integrate agents after 3 years
  • B2C model launched 1 month ago
  • Browser plugin planned with 90-day freemium period

Partnership Structure Discussion​

Bryan's Direct Question:

"How best we can go about this joint partnership, whatever kind of partnership it is, whether it's advisory, whatever it is, how best would you want us to go about it?"

Gary's Response:

  • Will put something down on paper
  • Needs to pray on it and discuss with Brenda
  • Considering "lesser co-founder" or "head of ecosystem" role
  • Emphasized symbolic empowerment through title

Timeline Urgency:

Gary: "It would be nice as I fly out to Rome on November 4th... to be able to represent DeepTrack whatever the capacity is in front of a bunch of delegates that care about faith and AI and represent big institutions."

Key Quotes​

On African Business Reality:

"The taxes are critically high, and then you also have to grease so many people along the pipeline. So if you would invest today for maybe $10,000, by the time it gets to you, it's probably 600 bucks."

On Product-Market Fit:

"By the time you convert this customer, it's probably a year... Maurice used to sell sofas for local governments here in Nairobi, in Africa. And he mentioned that there's a time that the only lead that he closed much faster was in three months."

On Technical Achievement:

"When I told him that one of our most successful use cases is insurance, he said that he still does not understand how we are able to do that, considering that is a very, very expensive entity."

On Divine Partnership:

Gary: "Probably getting laid off was the best thing that God could have ever done for you." Bryan: "Yeah. I think about it every single day. Truly, it was a blessing."

Action Items and Next Steps​

Immediate Actions​

  • Gary to draft partnership proposal within days
  • WhatsApp group chat to be created by Brenda
  • Follow-up call scheduled for Friday (November 1st) at same time
  • Gary to pray and consult with Brenda on structure

Strategic Priorities​

  • Formalize partnership before Rome conference (November 4, 2025)
  • Position for US market entry before December residency
  • Develop messaging to overcome "African scam" perception
  • Create enterprise sales materials for Western markets

Partnership Considerations​

  • Title options: Lesser co-founder, Head of Ecosystem, Strategic Advisor
  • Focus areas: US market development, fundraising, brand building
  • Vatican/Catholic Church integration opportunity
  • Developer ecosystem potential

Spiritual and Strategic Synthesis​

The conversation revealed a profound alignment between Gary's calling to elevate technical innovations and Bryan's mission to build truth infrastructure. Both men see God as central to the venture, with Bryan having independently declared God as co-founder before meeting Gary. The technical sophistication of building 26 specialized models with minimal resources, combined with Gary's proven ability to build movements around niche technologies, suggests divine orchestration. The timing - with Bryan's Silicon Valley residency approaching and Gary's Rome conference imminent - creates a strategic window for partnership that could transform DeepTrack from an African startup struggling with corruption and low returns into a global leader in AI trust infrastructure.