Section 01
The Problem: Traditional Due Diligence Fails in Web3
The standard institutional due diligence playbook was designed for private equity and public equities. It assumes audited financial statements, regulated disclosures, identifiable management teams with verifiable track records, and established legal frameworks. None of these assumptions hold in Web3.
Consider what an analyst faces when evaluating a liquid token protocol:
On-chain data is scattered across blockchains, layer 2s, and sidechains. Revenue is not reported in quarterly filings; it must be extracted from smart contract interactions. Team members may be pseudonymous or distributed across jurisdictions with no corporate entity. Tokenomics involve complex vesting schedules, inflation curves, and governance mechanisms that have no analogue in traditional finance. Community sentiment, developer activity, and ecosystem health are qualitative factors that require constant monitoring.
Traditional research teams attempt to manage this manually, but the scale is overwhelming. A single protocol can generate thousands of transactions per second. A portfolio of 15 protocols generates a firehose of data that no human can process systematically.
The result is that most institutional allocators in digital assets fall back on one of two approaches: they either delegate to external managers whose processes remain opaque, or they rely on narrative-driven decision making that is indistinguishable from retail speculation.
We built Wednesday to solve this problem.
Section 02
Wednesday: Our Proprietary AI-Native Investment System
Wednesday is an AI-native investment system purpose-built for digital asset research. It is not a single model or tool. It is an integrated pipeline that ingests, structures, and analyses data across multiple dimensions simultaneously.
The system operates on a continuous cycle. Every day, Wednesday crawls hundreds of sources including on-chain data providers, developer repositories, governance forums, social media channels, regulatory filings, and market data feeds. It processes raw data through a series of specialised models that extract structured signals from unstructured noise.
For on-chain analysis, Wednesday tracks key metrics across our coverage universe: total value locked, revenue, fee generation, user growth, transaction volumes, and token velocity. These metrics are not taken at face value. The system cross-references them against historical patterns, network benchmarks, and anomaly detection models to identify divergence from expected behaviour.
For qualitative analysis, Wednesday monitors developer activity on GitHub, proposal quality in governance forums, community health indicators, and team communication patterns. It analyses the substance of technical discussions rather than simply counting mentions. It evaluates whether development velocity is accelerating or declining, whether governance participation is broad or concentrated, and whether the community is aligned or fracturing.
For risk monitoring, Wednesday tracks wallet concentrations, large holder movements, exchange flows, smart contract interactions, and protocol-level vulnerabilities. It flags unusual patterns that warrant human investigation.
The output is not a single number or score. It is a structured dataset that our investment team uses as the foundation for every decision.
Section 03
PRISM: Our Deep Due Diligence Framework
Wednesday provides the data. PRISM provides the analytical framework. PRISM stands for Protocol Research, Investment, and Strategy Methodology. It is the structured process we apply to every potential investment.
PRISM operates across five dimensions:
Protocol Analysis. We evaluate the technical architecture, security model, and competitive positioning. Is the protocol solving a real problem? Is the design sound? Does it have a defensible moat? Wednesday provides the data; our team provides the technical judgment.
Tokenomics Assessment. We model the token's economic design including supply schedules, distribution mechanisms, incentive structures, and value accrual. We stress test these models under different adoption scenarios. We evaluate whether the tokenomics are aligned with long-term value creation or structured for short-term extraction.
Team and Governance Evaluation. We assess the team's track record, alignment, and operational transparency. For pseudonymous teams, we evaluate alternative signals such as code quality, community trust, and progressive decentralisation. We analyse governance structures to understand who holds power and how decisions are made.
Risk and Regulatory Mapping. We identify every material risk: smart contract risk, governance risk, regulatory risk, market structure risk, and liquidity risk. We evaluate jurisdictional exposure and legal standing. We do not assume risk away; we size it and decide whether we can manage it.
Validation and Cross-Referencing. We validate Wednesday's outputs against independent sources. We conduct our own on-chain analysis. We speak with teams, community members, and ecosystem participants. We challenge our own assumptions.
PRISM ensures that no investment decision is made on data alone. The system informs; the humans decide.
Section 04
AI as Input, Not Decision-Maker
A critical distinction must be made. Wednesday is an analytical tool, not an autonomous decision-maker. We do not use AI to generate trading signals or automate portfolio allocation. We use AI to structure information, surface anomalies, and compress weeks of research into hours.
The final decision always rests with our investment committee. Human judgment, experience, and governance overlay every output. This is not a limitation; it is a design choice.
Our risk management framework addresses this by maintaining a liquidity buffer, avoiding concentrated positions in low liquidity tokens, and using limit orders rather than market orders for execution. We also monitor on-chain exchange flows and perpetual futures funding rates as leading indicators of market stress. When funding rates turn deeply negative and exchange inflows spike, we reduce exposure preemptively.
AI systems in investment contexts suffer from well-documented failure modes: they hallucinate, they amplify training data biases, they fail in novel market conditions, and they lack the contextual understanding that comes from lived experience. We have seen funds that attempted fully automated approaches suffer catastrophic losses during market dislocations that their models had never encountered.
"Our approach is different. We treat AI as a force multiplier for human intelligence, not a replacement for it. Wednesday does the work that humans cannot do at scale. Our team does the work that AI cannot do at all."
n8 Capital Research
Section 05
How We Optimise LLMs Over Time
The landscape of large language models is evolving rapidly. Models that were state of the art six months ago are now obsolete. Most firms underestimate both the pace of this evolution and the difficulty of integrating new models effectively.
We take a systematic approach to model selection and optimisation. We maintain a continuous evaluation pipeline that tests new models against our specific use cases: on-chain data extraction, governance document analysis, developer activity interpretation, and risk signal detection. We do not assume that newer models are better. We test them rigorously against our benchmarks before integrating them.
Our system is designed to be model-agnostic. As better models emerge, we can swap them in without rebuilding the entire pipeline. This gives us a structural advantage in keeping pace with technological change without constant re-engineering.
Section 06
Why This Gives Us a Structural Edge
The combination of Wednesday and PRISM creates capabilities that no traditional fund can replicate easily.
Scale
A human research team of five people cannot monitor 15 protocols across 10 data dimensions in real time. Wednesday can. This means we see patterns and anomalies that our competitors miss.
Consistency
Human analysts have good days and bad days. They have biases, blind spots, and attention limits. Wednesday applies the same analytical framework every time, every day, without fatigue or bias. This produces more reliable signals over time.
Speed
When a market event occurs, we can assess its impact across our entire portfolio in minutes, not days. This allows us to make informed decisions during periods of volatility when others are still gathering information.
Learning
Every analysis, every decision, and every outcome feeds back into Wednesday's models. The system improves over time as it learns from our investment outcomes. Most funds have no equivalent feedback loop.
Section 07
Risks and Limitations
Wednesday operates within the limits of the data it can access. On-chain data is transparent but not always clean. Smart contract interactions can be obfuscated. Governance activity can be manipulated. The system can flag anomalies, but it cannot always distinguish between genuine signals and noise.
There is also the risk of over-reliance. A team that trusts its system too much may stop questioning its outputs. We guard against this through PRISM's validation layer and our investment committee's independent review. Every significant decision requires human judgment.
Finally, the system is only as good as the models that power it. We focus on ensuring we optimise the use of self-hosted (where possible) models in our orchestrated systems. These are synchronised to ensure we have the best information available to us to help us make the best decisions we can.
"At n8 Capital, we have invested heavily in building an AI-native system that matches the structural realities of the asset class. Wednesday provides the data. PRISM provides the framework. Our team provides the judgment. The combination gives us a scalable, consistent, and defensible edge.
n8 Capital Research
Section 08
Conclusion: The Institutional Standard
Digital asset investing demands a new approach to research. The old tools do not work. The data is too vast, too fragmented, and too fast-moving for traditional methods. Funds that rely on manual analysis will increasingly find themselves at a disadvantage.
We do not claim that our system eliminates risk. It does not. What it does is give us a clearer view of the risks we are taking, and a more systematic approach to managing them. For institutional allocators evaluating digital asset exposure, that clarity is invaluable.
n8 Capital advises family offices, high-net-worth investors, and institutions on digital asset allocation. Our approach combines proprietary AI-native research systems with disciplined risk management and institutional governance. We do not sell products. We provide expertise, analysis, and strategic guidance for sophisticated investors navigating the AI x blockchain convergence.