4
Active funds
How is your firm built?
Colibrí Architecture is the operating system for venture capital firm design. Take a short walkthrough. We read your firm across three independent dimensions, surface the tensions in how it is built, and produce a single research-grounded view of its architecture.
Step 01·Firm OverviewPractice
Your firm, read across every fund.
When you operate more than one fund, the questions your LPs ask are about the firm, not any single vehicle. Colibrí Architecture starts there. It reads every fund you operate side by side and gives you three scores in one place: an Architecture Score that combines all three analytical engines, a Cross-Fund Concentration Score, and a Firm CEI weighted by deployed capital.
Why this is Practice tier
The cross-fund view only makes sense if you have more than one fund. A single-fund firm sees the full analysis on its one vehicle through the Fund Overview tab. The Firm Overview tab, the Architecture Score, and Cross-Fund Concentration appear the moment you add a second fund.
01One integrated view.Architecture, concentration, and weighted efficiency in a single pane.
02Cohort positioning.Each score reads against the lifecycle cohort, not a generic benchmark.
Firm OverviewArchitectureFund OverviewFund DesignInvestmentsScenariosLP Memo
Firm Overview · Cross-fund analytical view
Firm Overview
4 Funds · $780M Committed · 52 Portfolio Companies
Architecture Score
Composite of three engines
91/100
Concentration
Firm-level distribution
95/100
Firm CEI
Capital-weighted efficiency
89/100
$780M
Committed
52
Portfolio cos
$24.0M
Avg GP load
Step 02·ArchitecturePractice
Three independent engines, one composite.
The Architecture Score combines three independent engines into one number. The math is set up so that a weak score on any one engine drags the whole score down. You cannot earn a strong Architecture Score through portfolio efficiency alone if your firm design is incoherent, and the reverse. That trade-off is structural, not a setting.
Why this is Practice tier
The composite needs all three engines, and the third one (Cross-Fund Concentration) only operates meaningfully when there is more than one fund to compare. So Foundation shows Engines 1 and 2 on the single fund. Practice shows all three engines plus the composite, which is the number most LPs look at first.
01Engine 1.Portfolio Efficiency. How well your capital deploys against your lifecycle cohort.
02Engine 2.Firm Design Congruence. Whether your structural choices fit together.
03Engine 3.Cross-Fund Concentration. Where the firm's capital is concentrated across funds.
04Composite.The integrated score. Imbalance brings it down by construction.
Firm OverviewArchitectureFund OverviewFund DesignInvestmentsScenariosLP Memo
Architecture · Firm analytical view
Architecture
Colibrí Ventures · Fund IV · 2024 Vintage
89 /100
Established · New
Engine 1
Portfolio Efficiency
87/100
5 subscores
Engine 2
Firm Design Congruence
85/100
11 checks, 3 soft
Engine 3
Cross-Fund Concentration
95/100
1 soft flag
Step 03·Fund DesignFoundation + Practice
The levers that determine everything else.
Every score in the platform traces back to a set of choices you make about how the fund is built: lifecycle stage, investment thesis, portfolio construction, decision process, partnership dynamics, support platform, and succession plan. Edit any lever and the rest of the platform updates. Eleven independent checks run continuously and surface tensions when two or more of your choices are pulling in opposite directions.
In both tiers
Fund Design is part of the core experience in both tiers. A single-fund Foundation customer gets the same eleven-check engine as a Practice customer running ten funds. The difference at this layer is how many funds it runs against, not how deep the analysis goes.
01Live congruence.The score updates as you adjust each lever.
02Named tensions.Each surfaced item names the variables in conflict.
Firm OverviewArchitectureFund OverviewFund DesignInvestmentsScenariosLP Memo
Fund Design · Configuration levers
Colibrí Ventures Fund IV
Series A-Focused · $400M Fund · 2024 Vintage
Investment Thesis02
Stage focus
Pre-seedSeedSeries AMulti-stageBroad
Industry scope
DeepFocusedThematicBroadGeneralist
Portfolio Construction03
$400M
20
15%
40%
Live Congruence ScoreModerately congruent
85/100
AlignedIC structure and industry scope. Voting model aligns with sector scope.
SoftStage focus and ownership. Ownership target sits just outside the expected band.
AlignedBoard seats and team capacity. Commitments fit team capacity at 1.7 seats per professional.
SoftLifecycle and succession. Succession work expected to have begun at this stage.
Step 04·Decision SupportFoundation + Practice
Every investment, evaluated against your configuration.
When you add an investment, the platform shows you how that one decision moves your portfolio before you commit to it. Stage discipline, sector spread, breadth, deployment pace, and check size relative to your benchmark all update as you type. It reads from the same configuration you set in Fund Design, so the analysis is consistent with everything else in the platform.
In both tiers
Decision Support runs on every investment, in every fund you operate. Foundation customers see how a new deal moves the one fund's portfolio. Practice customers also see how it changes the firm's total exposure if the company is one you have backed in another fund.
01Before commitment.The panel updates as you enter the investment, not after.
02Thesis stretch surfaced.The platform names when a deal sits outside the declared thesis.
Firm OverviewArchitectureFund OverviewFund DesignInvestmentsAdd InvestmentScenarios
New record · Configuration-aware
Add Investment
The Decision Support panel shows how this investment moves the portfolio.
Concord AI
AI/ML
Pre-seedSeedSeries ASeries BLater
$5M
$1.5M
8%
Decision Support · Live impact
Portfolio breadth6 / 20→7 / 20
Sector spread2→3
Stage discipline83% A→71% A
Deployment12%→13%
Modest thesis stretch
Seed investment falls outside the stated Series A focus. Adds AI/ML as a new sector. Check size $1.5M sits well below the benchmark check of $12.0M for this fund.
Step 05·Scenario EngineFoundation + Practice
Follow-on decisions become a structured output.
For each active position, the scenario engine asks one question: should you follow on, how much, and when? It returns an Outlier Probability Score (how strong the signals are that this company is heading for a power-law outcome), a recommended check size as a share of your pro-rata, a timing signal, and a three-part rationale. The recommendation curve is calibrated to your lifecycle: emerging managers see a steeper curve that rewards conviction, established firms see a flatter curve that favors measured diversification.
How the tiers differ here
Foundation ranks every active position in your one fund. Practice ranks positions across every fund in the firm, so the highest-conviction position in Fund III sits next to the highest-conviction position in Fund II. This matters when reserves are scarce and you need to put them behind the strongest signal, no matter which fund holds the position.
01Pro-rata as the anchor.Every recommendation is expressed as a share of your pro-rata, so the math is auditable.
02Reserve and concentration aware.Recommendations respect what is left in your reserve pool and your firm-level exposure limits.
Firm OverviewArchitectureFund OverviewFund DesignInvestmentsScenariosLP Memo
Scenario Engine · Follow-on analysis
Scenarios
Follow-on candidates across the firm, ranked by Outlier Probability Score.
Company · FundMOICOPSRecommendation
Brassroot
Fund II · Active
7.00×Yes · 1.4× pro-rataMarsh
Fund II · Active
6.00×Yes · 1.2× pro-rataHalden
Fund III · Active
4.50×Selective · 0.8× cappedCobble
Fund I · Partial exit
2.50×Wait · milestoneBramble
Fund I · Active
1.30×No · capital better elsewhereStep 06·LP MemoFoundation + Practice
The memo writes itself.
At the end of each quarter, the platform writes the LP memo for you. It pulls from every engine, traces every claim back to the configuration that produced it, and reads in plain institutional voice. The same data driving your internal dashboard becomes the document you share with your investors. No copying numbers between tools. No reformatting.
How the tiers differ here
Foundation writes a quarterly memo for your one fund. Practice writes a quarterly memo for each fund plus a firm-level annual letter, which is the report most institutional LPs actually read. Both tiers keep every prior memo with a snapshot of the scores at the time it was written.
01One source of truth.The same data drives the internal view and the LP-facing memo.
02Versioned history.Every quarter's memo is preserved with the scores as they read at the time.
Firm OverviewArchitectureFund OverviewFund DesignInvestmentsScenariosLP Memo
Quarterly LP Memo · Integrated Configuration Analysis · Q2 2026
Colibrí Ventures Fund IV
Portfolio and firm design analysis · Six months in fund · 12% of capital deployed
Fund IV closed its second full quarter of deployment with six initial investments totaling $49 million against the $400 million commitment. Stage discipline remained intact across the period, with approximately 83 percent of capital deployed at Series A, consistent with the stated thesis. The portfolio added one new sector position in the period and now reads as a three-sector focused mandate.
The firm's Architecture Score for the quarter is 89, positioned in the top quartile of established firms by lifecycle cohort. The three engines that compose this view returned a Composite Efficiency Index of 87, a Configuration Congruence of 85, and a Cross-Fund Concentration Score of 95. The three soft tensions surfaced in the quarter are documented in the appendix.
89
Architecture
$49M
Deployed
6 / 20
Portcos
Step 07·MethodologyFoundation + Practice
Every number is auditable.
The methodology document specifies how every score is computed, end to end. Three engines, one composite, a temporal layer that tracks how your scores change over time, the scenario engine, and a published process for how the calibration evolves. No black boxes. Your LPs can audit the framework without needing access to our source code.
In both tiers
The methodology document is the same for both tiers. The difference is which sections apply to your view. Foundation operates against four sections (Engines 1 and 2, the Temporal Layer, and the Scenario Engine). Practice adds two more on top: Cross-Fund Concentration and the Architecture Score.
01Published thresholds.Every band, weight, and threshold is documented and versioned.
02Calibration governance.Updates to the framework follow a documented review process, never quietly.
Platform Methodology · Version 0.3 · Working Draft
Colibrí Architecture
IFoundations
IIData Architecture
IIIEngine 1 · Portfolio Efficiency
IVEngine 2 · Firm Design Congruence
VEngine 3 · Cross-Fund Concentration
VIThe Architecture Score
VIIThe Temporal Layer
VIIICalibration Governance
IXScenario Engine
End of tour · Choose your scope
Two ways to operate.
Colibrí Architecture comes in two tiers, sized to the structural complexity of the firm using it. Foundation is built for a partnership operating a single fund. Practice is built for partnerships operating more than one. The methodology is the same in both. What changes is the scope of the analysis.
Tier 1
Foundation
For partnerships operating a single fund. The complete configuration analysis and decision-support engine for the vehicle you are running today.
What you get
- Portfolio Efficiency analysis.How efficiently your capital deploys, scored against firms at your lifecycle stage.
- Firm Design Congruence.Eleven checks that surface where your structural choices conflict with each other.
- Live Fund Design.Edit any lever and watch the scores update across the platform.
- Decision Support on every investment.See how each new deal moves the portfolio before you commit.
- Follow-on scenarios.A recommended check size and timing signal for each active position.
- Quarterly LP memo.Generated from your dashboard, ready to share with your investors.
Tier 2
Multi-fund unlockPractice
For partnerships operating two or more funds, up to ten. The cross-fund view your institutional LPs are actually asking about.
Everything in Foundation, plus the firm-level analysis that only exists when you operate more than one fund:
Practice adds
- Firm Overview.All your funds, side by side, sorted by vintage.
- Cross-Fund Concentration.Where the firm's capital is concentrated across sectors, companies, and stages.
- Architecture Score.The integrated score that brings all three engines into a single number.
- Weighted Firm CEI.Portfolio efficiency across every fund, weighted by deployed capital.
- Firm-wide follow-on ranking.Compare every active position across all funds, ranked by conviction signal.
- Firm-level annual LP letter.The consolidated cross-fund report, in addition to the per-fund quarterly memos.
Three independent engines. One composite.Eleven firm-design checks. Continuously evaluated.Firm-private data. Sovereign at the firm boundary.