Based AI-Driven Knowledge Services - Operating Model

[Timestamp: 26_0506 23:35:00] [Session-ID: 18f92619-bd24-4aa7-8681-c15d3122c23f] [Runtime: Copilot_CLI] [Agent: financial_analyst]

Objective

Translate the MGA14 knowledge-tech architecture into a Based operating model for repeatable customer-facing knowledge services.

Operating spine

```text

Source registry

-> inbound capture

-> raw asset log

-> markdown asset log

-> HROI gate

-> accepted asset cluster

-> cognitive links and persona graph

-> NotebookLM / vector-ready knowledge base

-> customer outputs

-> memory, lessons, and refresh loop

```

07 May HROI revenue-floor update

The operating model now needs an explicit revenue-source separation:

Revenue layer	Operating treatment	Gate
Current MSC/MSB Breakfast floor	Treat EUR 0.5m as prior-year generated event revenue expected to repeat this year.	Source/accounting tie-out before investor release.
New intelligence products	Track Intelligence Room, Watchdesk, Diligence Sprint, sponsor/member intelligence, and strategic-program revenue as incremental expansion above the floor.	Buyer, product, price, volume, source, evidence state, owner.
Expansion ceiling	Treat EUR 2.5m as 2026 total ceiling including the current floor.	Named pipeline, pilot conversion, delivery capacity, and margin proof.
Legacy high scenario	Keep as diligence appendix only.	Not first-send base case.

Required components

Component	Function	Based product role
Source registry	Defines approved source classes and access posture.	Prevents unreliable or rights-unclear inputs.
Raw asset capture	Logs source, date, author, path, and capture metadata.	Creates audit trail.
Markdown normalization	Converts materials into readable, linkable knowledge assets.	Makes outputs reusable across briefs, rooms, and dashboards.
HROI quality gate	Scores provenance, relevance, novelty, credibility, actionability, durability, cognitive-link value, and agent-skill value.	Trust layer and product differentiator.
Cognitive linking	Links assets to adjacent assets, people, companies, domains, and research questions.	Compounds insight over time.
Persona graph	Models investors, founders, analysts, regulators, institutions, companies, and sectors.	Supports stakeholder strategy and influence mapping.
Training asset workflow	Converts accepted assets into briefing notes, question banks, and decision memos.	Turns intelligence into capability.
Research orchestration	Runs recurring source pulls, watchlists, source-gap reports, and briefs.	Creates subscription cadence.
Directory hygiene	Keeps duplicates, stale assets, orphan rows, and unsafe connectors out of the system.	Maintains reliability as volume grows.

Curator domains for Based

Domain curator	Based use
Frontier green technology	Climate adaptation, carbon, materials, circularity, water, biodiversity technology.
Frontier energy technology	Storage, grid, nuclear, hydrogen, geothermal, power electronics, energy-market infrastructure.
Global tech markets	AI infrastructure, developer ecosystems, platform shifts, enterprise technology.
VC/PE markets	Funding conditions, valuation signals, exit environment, investor strategy.
Capital markets	Rates, funding markets, fund flows, public-market context for resilience capital.
Hot early-stage companies	Startup watchlists and emerging resilience company maps.
European regulation	EU policy and regulatory change relevant to infrastructure, climate, energy, and data.
Defense-tech / industry / regulation	Optional adjacency where resilience, security, infrastructure, and public procurement overlap.

HROI gate for customer-facing intelligence

Every asset promoted into customer-facing output should include:

HROI score;
HROI state: `accepted`, `review`, or `rejected`;
hard blocks;
decision axis;
cut;
knowledge state;
generated-by stamp;
review gate owner;
next route;
cognitive links;
skill-ingest candidate flag;
training-asset candidate flag.

Customer output factory

Input pattern	Output
High-HROI source cluster	Monthly sector heartbeat.
High-impact company signal	Company watchlist note.
Regulation or public funding signal	Regulation change note.
Investor/founder/analyst cluster	Persona and influence graph update.
Repeated customer question	Training asset or decision memo.
Evidence-rich opportunity area	Diligence sprint pack.
Source coverage gap	Research backlog and source-gap report.

Review gates

Gate	Rule
Provenance	No customer-facing use without clear source, date, author, capture route, and access note.
Financial/regulatory interpretation	Label as framing unless jurisdiction-specific advice is separately substantiated.
Ranking or prioritization	Must include influence axis and cut.
Frozen knowledge	Must include caveat, generated-by stamp, review gate, and refresh cadence.
Connector expansion	Requires safety review before live automation.
Customer/investor promotion	Requires tier decision and approval evidence.

Launch operating cadence

Cadence	Activity
Weekly	Source capture, HROI triage, accepted asset cluster, signal brief.
Monthly	Sector heartbeat, watchlist update, persona graph update, source gap report.
Quarterly	Influence-axis refresh, domain cut review, pricing/package review, customer value review.

Minimum viable pilot

The smallest credible pilot should include:

1. one source registry;

2. one Based Resilience Intelligence Room;

3. one HROI-scored source index;

4. one sector heartbeat report;

5. one company/technology watchlist;

6. one investor/persona graph;

7. one training or decision asset;

8. one source-gap report.

Operating decision

Start with a managed-service operating model and keep automation behind review gates. The product can later expose self-serve dashboards, but early customer trust depends on curated outputs, clear provenance, and visible review discipline.