AI can analyze a pitch deck in seconds. It can scan markets, benchmark metrics, and flag inconsistencies faster than any human.
But AI can't call the founder's former colleagues. It can't sense the subtle signals of a team that's about to implode. It can't draw on twenty years of pattern recognition from backing (and missing) breakout companies.
That's why Athanor combines AI analysis with structured human expertise. Here's exactly how it works.
The Two Layers of Human Input
1. Circle of Alchemists (CoA)
The CoA is our core team—the people who built Athanor and live in dealflow every day. Right now, that's three of us.
Who we are:
- Founders who've been on both sides of the table
- Operators who've built, scaled, and sometimes failed
- People who read pitch decks for breakfast
What we do:
- Review GemScore reports for calibration
- Add qualitative context AI can't capture
- Flag pattern-match signals (positive and negative)
- Provide "would I take the meeting?" gut checks
How we expand: When a deal needs domain expertise we don't have, we reach out to our network—case by case. We don't maintain a bench of advisors waiting for assignments. We identify the right person for the specific question, make the introduction, and bring them in for that engagement.
2. Invited Domain Experts
When a deal requires specialized knowledge—deep tech validation, regulatory insight, market-specific context—we bring in experts from our network.
Who they are:
- Technical experts (CTOs, engineers with deep domain knowledge)
- Industry veterans (former executives in the target market)
- Regulatory specialists (for healthcare, fintech, etc.)
- Geographic experts (for market-specific context)
What they do:
- Provide focused reviews on their area of expertise
- Answer specific questions flagged by AI or CoA
- Validate (or challenge) technical claims
- Assess competitive positioning from insider perspective
How it works: This isn't a marketplace with open applications. We work with the best experts we know—people we've built relationships with, who we trust to give honest assessments. When we need someone, we reach out directly. Quality over scale.
Experts can charge for their time and leverage their knowledge. We handle the coordination and structured output collection.
The Review Flow
Here's how a typical evaluation moves through our system:
Stage 1: AI Analysis (GemScore)
The founder submits their idea. Our AI engine:
- Analyzes the pitch across 30+ dimensions
- Pulls market data, competitive landscape, team backgrounds
- Generates a structured GemScore report
- Flags areas that need human review
Output: A comprehensive report with scores, evidence chains, and identified gaps.
Stage 2: CoA Calibration
For ideas that warrant deeper review, the report goes to a matched CoA member:
- They review the AI analysis
- Add qualitative notes and pattern-match signals
- Adjust scores if they see something AI missed
- Flag anything that needs expert input
Output: Calibrated report with human annotations.
Stage 3: Expert Deep-Dives (Optional)
If specific questions remain—technical feasibility, regulatory risk, market nuance—an invited expert weighs in:
- They receive targeted questions (not the full report)
- Provide focused analysis in their domain
- Their input is integrated into the final assessment
Output: Expert opinions attached to specific report sections.
Stage 4: Final Synthesis
The complete picture is assembled:
- AI foundation + CoA calibration + expert insights
- Clear summary of strengths, concerns, and open questions
- Recommended next steps (pass, dig deeper, take the meeting)
Why This Model Works
For Investors
- Leverage: Get expert-level analysis without expert-level headcount
- Consistency: Every deal gets the same structured treatment
- Speed: AI does the heavy lifting; humans focus on judgment calls
- Customization: Bring in your own experts when you need them
For Invited Experts
- Monetization: Get paid for your domain knowledge
- Flexibility: Engage on specific deals, not ongoing commitments
- Reputation: Build a track record of valuable insights
- Access: See interesting dealflow through a trusted channel
For Founders
- Quality feedback: Not just a yes/no, but structured reasoning
- Multiple perspectives: AI + generalist + specialist viewpoints
- Credibility: A report backed by named experts carries weight
- Actionable: Clear on what to improve, not just what's wrong
Skin in the Game: Accuracy Tracking
Here's where it gets interesting.
We track advisor and expert accuracy over time. When they make predictions or assessments, we record them. When outcomes become known—company raises next round, hits milestones, fails—we update their accuracy scores.
This creates:
- Calibration: Experts learn where their judgment is strong (and where it's not)
- Signal: Investors can weight opinions by track record
- Incentives: Better predictions = more review opportunities = more compensation
It's not about being right every time. It's about knowing your hit rate and being honest about confidence levels.
The Tokenized Promise Connection
Our work on Tokenized Promises ties directly to this expert model.
Imagine an expert who says: "I believe this company will hit $1M ARR within 12 months."
With a Tokenized Promise, that opinion can become a commitment:
- Expert stakes reputation (and optionally, capital) on the prediction
- If correct, they earn premium compensation
- If wrong, stake is reduced
This is still experimental, but it points toward a future where expert opinions aren't just words—they're on-chain commitments with real consequences.
How to Access Human Reviews
For investors on Athanor:
- CoA review is included for qualified deals
- Need specialized expertise? We'll tap our network
- Bring-your-own-expert option available (contact us)
For experts: We don't have an open application process. We work with people we know—through direct connections, mutual relationships, and proven track records. If you're exceptional at what you do and we cross paths, we'll reach out when there's a fit.
For founders:
- Request expert feedback when submitting ideas
- See which experts reviewed your pitch (with their permission)
- Use expert validation in your fundraising materials
The Vision
Today, we're small and selective. The CoA is our core team, expanded case-by-case with trusted experts from our network. That's intentional—quality scales better than quantity at this stage.
Over time, we're building toward:
- Every founder gets structured, actionable feedback
- Every investor has on-demand access to expert judgment
- Trusted experts can monetize their pattern recognition through our network
- Every opinion is tracked, calibrated, and optionally committed on-chain
AI handles the scalable parts. Humans handle the judgment calls. The combination beats either alone.
Want to see how GemScore + human review works on your dealflow? Let's talk.
