Signal Chat Synthesis: Jan 14 – Apr 11, 2026

Synthesized from 2,016 substantive messages. This document surfaces patterns, contradictions, gaps, and open questions — it does not draw final conclusions. Conclusions belong to Bliss and Dustin.

The Arc of Thinking

Jan – Early Feb: Broad Exploration Mode

The chat opens in exploration mode. Four strategic buckets are in play simultaneously; the founders are scanning broadly before narrowing. The oil-analog research sprint is a structuring device — using well-understood commodity market dynamics as a lens for understanding where supply chain intelligence products have succeeded and failed.

First concrete customer signal arrives Feb 7: Eoin O’Doherty, flagged in real time as “absolute gold” — an operator describing an actual compliance pain point in specific operational terms. The advisor network begins forming: Iancu, Dasher, Blank are each engaged for the first time or newly activated.

Mid Feb: Batteries Pivot and Narrative Discipline

The oil-analog scan returns a surprising output: batteries, not semiconductors, have a clearer near-term regulatory forcing function (EU Battery Passport). The founders seriously consider pivoting the Day-1 wedge. This is the most consequential inflection point in the chat.

The Botha Chan application runs in parallel, forcing narrative compression. “TierZero” emerges as the submission name. The application process functions as thesis discipline — you can’t be vague when you have a word limit.

The Steve Blank meeting on Feb 18 (“we fucking crushed that”) becomes a milestone. The three-step vision — compliance → intelligence → finance — is locked as the canonical frame. Blank’s affirmation appears to stabilize the founders’ confidence at a moment when the batteries/semis question is still open.

Late Feb – Early Mar: Crystallization

The petal diagram synthesis produces the most important intellectual artifact of the period. The thesis sharpens into a specific causal chain: compliance workflows generate proprietary data exhaust; data exhaust powers a digital twin; the digital twin enables financial products. “Data exhaust” is adopted as a term of art.

Two competitive insights crystallize in this period. First: Altana and Interos have broken software — the market is not well-served by incumbents despite real demand. Second (from Eyck): the founders’ apparent weakness (no customer insight yet) is actually a strength if positioned correctly — the compliance wedge gets inside the customer organization and generates data incumbents can’t access.

Mid Mar: Null Hypothesis Locked on Semiconductors

Battery pivot is shelved. The null hypothesis returns explicitly to semiconductors. The decision is framed as conditional — “unless shown evidence otherwise” — but the conversation moves on quickly.

The McMaster meeting yields a commercial insight that doesn’t appear anywhere in prior documents: cost engineers, not compliance teams, hold the relevant budget. The product framing shifts from binary certification to probabilistic modeling. Neither of these insights has been integrated into public materials as of Apr 11.

Botha Chan selection confirmed. Eleanor Lacey surfaces unexpectedly as a high-value connector — a former Asana GC met during spring break travel who offers introductions into the GC network.

Apr: Execution Mode

The chat transitions to tactical coordination. Q2 goals and the April Tactical Plan are produced. Ontology building is identified as the most critical near-term technical task. Technical infrastructure (Claude Code + Eitan) is being stood up.

Steve Blank cancels his Apr 7 meeting. The founders treat it as data rather than a setback — a measured response. Re-engagement is planned via the tactical plan document.

Recurring Themes

1. Speed vs. Depth Tension

Dustin consistently pushed for faster external engagement and iteration. Bliss pushed for deeper research and epistemic rigor before locking positions or codifying outputs. The dynamic was productive across the period: it generated both forward movement and quality control. The February research sprint (Dustin rearranged his calendar to protect it) was a resolution in Bliss’s direction; the March acceleration of external meetings was a resolution in Dustin’s.

Worth watching in Q2: if this tension becomes a recurring negotiation rather than a productive creative friction, it may slow execution without generating new learning.

2. Programs as Tools, Not Goals

H4D, Botha Chan, RDI, and the LLP structure were each evaluated consistently through the lens of “does this serve our business objective?” The founders showed consistent resistance to letting any program or framework define the direction rather than serve it. This discipline was explicit — articulated out loud, not just practiced.

Strong alignment between founders on this throughout the period.

3. AI-Native as Competitive Advantage

The theme appears in three distinct registers across the arc: as intellectual excitement (Jan 21, discussion of what AI-native companies can do), as product principle (Feb 21: “AI-native companies will out-compete legacy software”), and as execution tool (Apr: Claude Code, rig setup, Eitan engagement). The progression is coherent and consistent.

What’s less clear: whether “AI-native” has been tested as a value proposition with actual customers. It is asserted as a differentiator but the customer-side evidence base is thin.

4. National Interest as Mission Frame

Both founders return repeatedly to connecting Project TBD to US national security, geopolitical competition, and the “collective West.” The framing is authentic — rooted in Bliss’s Palantir background and Dustin’s Ukraine supply chain experience — and is likely load-bearing for certain fundraising conversations (defense-adjacent investors, government relations).

The question worth sitting with: is national interest framing load-bearing for the commercial product, or is it a parallel motivational layer? If the buyers are cost engineers at semiconductor manufacturers (per the McMaster insight), how much does the national security frame matter to them vs. the efficiency/risk frame?

5. Advisor Credibility Cascades

The advisor network is building through compounding referrals: Steve Blank’s name opened conversations with Ann Miura-Ko and Jillian Manus. Holland Ferguson brokered introductions to Felter and Porteous. Eyck provided contacts. McMaster introduced Singleton and Baxter. Each node unlocks more nodes.

This is a strong pattern. The risk is that the advisor-to-customer ratio remains high — advisors are not substitutes for customers, and the RDI methodology prioritizes customer learning above expert validation.

6. Thesis Oscillation Before Locking

The core thesis moved: semiconductors → batteries → semiconductors across Feb–Mar. Both movements were data-driven (oil-analog scan recommended batteries; existing synthesis docs and advisor alignment pulled back to semis). The founders treated this as scientific updating, not inconsistency — a reasonable framing.

The underlying question it raises: is the semiconductor conviction now based on deeper research conducted between Feb 17 and Mar 18, or on accumulated path dependence (the business was built around semis, the name “TierZero” implies it, the advisors know it)? The distinction matters for how firmly the null hypothesis should be held.

Decisions Made But Never Formally Documented

These decisions were made in Signal conversation but do not appear to have been written into any strategy document, working agreement, or process doc in the vault:

3-business-day follow-up cadence (Jan 22) — Standard practice for all outreach. Not reflected in any documented workflow or the Outreach Style Guide.
Founders finish their own conversation threads (Jan 20) — Working norm about outreach ownership and relationship integrity. Not documented anywhere.
Divergent AI research queries as epistemic discipline (Feb 5) — The practice of not sharing prompts before comparing outputs, to preserve independent analysis. Not documented.
Don’t pitch financial products too early (Ann Miura-Ko, Mar 6) — Her specific advice: “Focus on compliance data collection now, worry about derivatives/insurance later.” Appears in her interview file but has not been integrated into any strategy document as a guiding constraint.
Cost engineers as primary buyer persona (Mar 18) — The most commercially significant insight from the Joe/Ali meeting. Not yet reflected in the OnePager, pitch materials, or any persona document.
Probabilistic model, not binary certification (Mar 18) — A product framing shift with significant implications for what the MVP looks like. Not yet integrated into public materials.
Trojan horse as core differentiator (Apr 6) — Explicit acknowledgment that compliance-as-data-acquisition is the strategic edge, not the compliance product itself. Partially articulated in the Charter but not consistently foregrounded across materials.

People Mentioned Repeatedly as Important

Richard Dasher (Stanford APARC)

Appeared across the full arc as a core recurring advisor. Met Jan 15, Feb 25, Apr 7. Sent four core artifacts. Offered an introduction to a Mitsui visiting scholar that has not been followed up as of Apr 11. Likely has more to offer on Asia supply chain dynamics given his APARC role.

Steve Blank

Anchor advisor and credibility anchor for the advisor network. The Feb 18 meeting (“we fucking crushed it”) was a milestone. The Botha Chan and Deep Dive #2 Charter were organized partly around preparing for him. Apr 7 meeting canceled — founders treated it as data. Re-engagement via tactical plan document is planned but not yet complete.

Holland Ferguson

Proactively brokered introductions (Felter, Porteous) without being asked. Described in the chat as a “weapon” and a future operator candidate. Status: Isobel Porteous introduction offered but not yet executed. The relationship appears warm but is not being actively advanced.

Eoin O’Doherty

First organic customer signal of the entire project (Feb 7). Described by founders in real time as “absolute gold” — an operator articulating a specific compliance pain point. No formal interview has been documented. Status as of Apr 11: unresolved.

Daleep Singh

Gave an unprompted policy-level validation in a class setting (Feb 17) — described as aligning precisely with the founders’ thesis on supply chain as a geopolitical instrument. Flagged as high-priority outreach. Status: zero documented follow-up.

Eleanor Lacey

Former Asana General Counsel. Met unexpectedly during spring break travel. Offered introductions into GC networks at large tech companies — potentially a direct path to compliance buyer personas. Multiple follow-ups planned (dinner, GC intros) but not yet executed as of Apr 11.

Ali Keshavarzi

Described in the chat as the most tactically useful meeting of the Mar 18 cluster. Offered introductions to Microsoft Chief Economist, a former McKinsey partner, and an Exxon contact. Follow-up meeting requested by the founders but not scheduled as of Apr 11.

Contradictions

1. Batteries vs. Semiconductors

Feb 17: The oil-analog scan — the founders’ own commissioned research — recommended batteries as the highest-conviction Day-1 wedge, citing the EU Battery Passport as a clearer near-term regulatory forcing function than anything in the semiconductor compliance stack. Mar 18: null hypothesis is locked back on semiconductors.

The reversion happened and was accepted. What’s less clear from the Signal record: what new information or reasoning drove the return? The chat references “existing synthesis docs” and advisor alignment, but doesn’t document the specific evidence that shifted the balance. This suggests the reversal may have been driven more by comfort and momentum than by new data.

2. Research Sprint Integrity vs. External Meeting Pace

Jan 29: Dustin rearranged his entire February calendar to protect an internal research sprint — a strong commitment to depth before external engagement. Mar–Apr: external meetings accelerated significantly (McMaster, Joe/Ali, Dasher, Blank, Eyck) in close succession. The question is whether the transition from sprint mode to external mode was deliberate and timed, or whether external meeting opportunities simply accumulated faster than the sprint outputs were processed.

3. Blank vs. Miura-Ko on Pacing

Steve Blank’s guidance (as reflected in how the founders prepared for him): ground everything in deep customer discovery before building. Ann Miura-Ko’s guidance (Mar 6, explicit): move faster, focus on data collection now, the financial products can come later. These are not the same advice. The founders noted the divergence in the chat but did not appear to explicitly resolve which approach governs Q2 execution.

4. “We haven’t gone deep enough on semis” vs. “Null hypothesis is semis”

Feb 15 (Bliss, explicit): “We haven’t gone deep enough into semis.” Mar 18: the null hypothesis is locked on semiconductors. The depth work between those two moments is not clearly surfaced in Signal. What changed? What did they learn? The answer matters for how firmly the semiconductor conviction should be held going into Q2 customer conversations.

Based on what the Project TBD architecture and RDI methodology say should be covered, these areas appear underweighted relative to their stated importance:

1. Actual Customer Interviews

The RDI methodology calls for deep customer conversations as the primary epistemic source. The most concrete customer signal in the entire 3-month period is Eoin O’Doherty (Feb 7), who was flagged as gold and never formally followed up. The formal interview pipeline is significantly thinner than the advisor pipeline. Advisors validate; customers reveal. The imbalance may be optimizing for confidence rather than learning.

2. Operators in the Compliance Workflow

The methodology explicitly calls for “operators — people actually doing compliance day-to-day.” The advisor conversations dominate the record — strategists, investors, policymakers, academics. Very few people who actually sit in a compliance role at a semiconductor company appear in the interview record. The McMaster meeting (Ali, a cost engineer) is the closest analog. This is the most significant gap relative to the RDI framework.

3. Competitor Analysis Evidence Base

Altana and Interos are described as having “broken software” — an important competitive insight. Wirescreen and Exeger appear in Eyck’s notes. But the competitive landscape summary (memory/media/2026-03-01-landscape-summary.docx) hasn’t been synthesized into the vault in a queryable form, and the “broken software” claim appears to rest on the founders’ observation rather than customer interviews. What do actual users of Altana or Interos say about them?

4. The Buyer Side of Financial Products

The three-step vision ends with parametric insurance / financial derivatives on semiconductor supply chain risk. This is the business model justification for everything upstream. But who buys these products? The chat is light on counterparties: insurance carriers, commodity desks, hedge funds, corporate treasury functions? The Munich Re / battery company model is referenced in the Charter but not explored in Signal. No conversations with financial risk buyers appear in the interview record.

5. Direct Regulatory Engagement

The RDI methodology calls for talking to regulators: “What is enforcement actually focused on? What are they worried about?” Eyck’s perspective on regulation is the closest thing in the record. No direct conversations with BIS, CBP (UFLPA enforcement), or OFAC are documented. Regulatory conversations could reveal enforcement priorities, data needs, and partnership opportunities that are invisible from the private sector side.

6. International Perspectives

The semiconductor supply chain is structurally international — Taiwan, South Korea, Japan, the Netherlands are central nodes. Bliss was in Seoul in late March and photographed the Samsung campus. Dustin asked about a possible Hong Kong trip. But apart from Richard Dasher’s Asia expertise, international practitioner perspectives are almost entirely absent from the interview record. This is particularly significant given the thesis involves mapping global supply chain relationships.

Open Questions for Bliss and Dustin

Eoin O’Doherty gave you the most concrete customer problem statement in the entire three-month period. You flagged it as “absolute gold” in real time. What happened to that thread? Is he still reachable? What would it take to do a formal interview now?
The batteries pivot (Feb 17) and the return to semiconductors (Mar 18) both happened in less than a month. What specifically drove the return? Is the semiconductor conviction now resting on deeper research conducted in that window, or on path dependence and advisor comfort with the original framing?
Ali Keshavarzi’s insight about cost engineers as the buyer persona (Mar 18) is the most commercially significant reframe in the chat — and it contradicts the earlier assumption of compliance teams as buyers. Has this been pressure-tested with anyone else? Should the OnePager be updated before the next round of investor conversations?
Daleep Singh gave you an unprompted policy-level validation in a class setting and you flagged him as a high-priority outreach target. That was in February. What’s the path to that conversation? What’s the barrier?
Holland Ferguson has been described as a “weapon” and a future operator candidate. What would need to be true for you to bring her in formally? What does the relationship look like if you don’t activate her — is it being maintained or fading?
You have a cluster of high-priority people whose follow-ups went unexecuted between March and April: Eoin O’Doherty, Daleep Singh, Eleanor Lacey, Ali Keshavarzi. Which three should be at the top of the outreach queue right now, and what’s the specific ask for each?
The thesis endpoint is parametric insurance or financial derivatives on semiconductor supply chain risk. You have not, as far as the Signal record shows, had a single conversation with anyone on the buyer side of those products — an insurance carrier, a commodity desk, a corporate treasury team. What would it take to get a Munich Re or similar counterparty into a conversation before the end of Q2?

Synthesis prepared from Signal chat export (2,026 messages, Jan 14 – Apr 11, 2026). Sourced: signal-import. Last updated: 2026-04-14.

Project TBD Memory Vault

Explorer

signal-synthesis