Clarenza is an agentic due diligence platform built for private capital. It reads an entire data room (VDR exports, pitch decks, CIMs, financial models, contracts) and turns it into IC-ready investment artifacts: screening memos, KPI tables, DDQ answers and a Red Team review. Every figure links back to its exact source, every calculation is computed deterministically rather than guessed by a language model, and every correction your analysts make is kept in a private memory that makes the platform sharper with each deal. The result is faster conviction on more deals, without giving up accuracy or control.

European private capital teams: PE buyout funds, venture and growth equity investors, private credit, corporate venture arms and family offices making direct investments. It is designed for firms where each analyst handles a high volume of document-heavy deals and where every hour saved on mechanical work goes back into judgment. Dedicated workflows exist for each fund archetype, and the platform is configured to your own fund thesis rather than a generic template.

Both, along one workflow. Most teams start with screening: a first memo on a new opportunity in under 30 minutes, scored against the fund's hard gates and criteria. As a deal progresses, the same deal record deepens into due diligence: full data room coverage, KPI extraction and reconciliation, questionnaire answers, Red Team review and the audit trail that goes to IC. The memo is a living artifact that evolves with the deal rather than a one-off document.

General assistants are excellent at drafting and explaining. Investment decisions ask for something else: numbers you can trust, evidence you can open, and a record your firm can stand behind. Clarenza differs in five ways:
‍
- Computed, not guessed. KPIs are calculated by deterministic solvers, never by the language model. Each calculated field shows its formula and inputs, and gaps between reported and recomputed figures are flagged.
- Traceable. Every figure links to the exact page, cell, or slide it came from, with the passage highlighted. A general assistant may cite a source but cannot reliably pin a figure to a document version or keep that link for audit.
- Consistent. The whole team works from one structured, verified data set, so the same question gives the same answer. Chat outputs vary across runs, users, prompt wording, and silent model updates.
- Complete. Clarenza processes the whole data room once, every page and sheet. Chat tools are bound by per-project file caps and retrieval quality, and accuracy drops as context grows.
- Built for a firm. Fund thesis configuration, DDQ workflows, deal-level permissions, institutional memory, approval routing, and an audit trail. None of these exist in a chat window. And on data, your deals never reach a US model provider, are never used for training, and stay inside EU infrastructure under your firm's control.

And much more! You can learn more here.

Three architectural choices. First, radical source traceability: every metric links to its exact location in the original document through our patent-pending Evidence Previewer, so verification takes one click instead of a search through the data room. Second, deterministic reasoning: financial calculations run on a neuro-symbolic engine with custom-built financial solvers, which means less reliance on frontier models and no hallucinated arithmetic. Third, institutional memory: every analyst correction, override and decision compounds into a private, tenant-isolated knowledge base that your whole team can draw on. Around these, Clarenza adds collaboration features built for how deal teams actually reach consensus, and a sovereign EU deployment that passes compliance review.

An IC-ready memo in your firm's own template, typically within 30 minutes of uploading a data room. It includes extracted and normalized financials, computed KPIs with their formulas, a Red Team section with kill conditions, flagged discrepancies and anomalies across sources, and a list of open questions for the company. Alongside the memo you get a prefilled DDQ, a searchable deal record, and the full chain of agent reasoning behind every conclusion.

Anything a data room contains: PDFs, pitch decks, CIMs, Excel financial models, contracts, cap tables, management presentations and board reports, including scanned documents. Clarenza ingests unstructured and semi-structured data at the page, cell and slide level and extracts both claims and metrics with their coordinates in the source. Beyond the data room, it enriches deals with market data from providers such as PitchBook and Crunchbase and can pull context from your CRM, email and meeting notes through connectors.

Yes. Clarenza processes the entire room once, every page and every sheet, into structured intelligence that our agents can query repeatedly. There is no per-project file cap and no sampling. When the data room changes, new or updated documents are reconciled against the existing record and the changes are tracked across versions, so you always know which figure came from which version of which file.

That is the point of the architecture. Figures are extracted, validated and linked to their source before they become trusted data. KPIs such as growth rates, margins, multiples and unit economics are computed by deterministic solvers, never by the language model, and each one shows its formula and inputs. When a reported figure and a recomputed one disagree, or when two documents contradict each other, Clarenza flags the discrepancy instead of picking one silently. A wrong multiple should never look like a right one.

Every figure in a memo, table or DDQ answer is clickable. The Evidence Previewer opens the original document at the exact page, cell or slide, with the relevant passage highlighted. You see the source, the extraction and any calculation applied, and you can correct or override any of it. Corrections are stored in your firm's private memory and applied going forward.

Yes. During onboarding you configure your fund thesis: geography and sector focus, stage and ticket size, growth and profitability expectations, ownership and deal type, fund-level constraints and return targets. Each criterion is set as a hard gate, a scored factor or context. Every memo is then written against that thesis, so a buyout fund, a growth investor and a venture fund each get different screening logic from the same data room. Firms with several funds can configure a thesis per fund, and theses are templated, refined and versioned over time.

An adversarial pass that stress-tests every memo before it reaches your team. Red Team agents argue the counter-case, check the deal against your fund's kill conditions, and surface discrepancies, anomalies and gaps across the data room. General assistants tend to follow the framing they are given; Red Team makes sure the case against the deal is on the table whether or not someone remembers to ask for it.

Yes. Import your firm's DDQ and Clarenza prefills it from the data room, with every answer sourced to the evidence. It also proposes deal-specific follow-up questions based on what the documents do and do not say, so your request list to the company is ready in the same pass.

Deals live in one shared workspace. Memos, tables and DDQs are shared artifacts with comments, mentions, tags, templates and firm-wide settings, and open questions can be assigned and tracked to resolution. Everyone works from the same verified data, so conviction is built together on one fact base rather than in parallel documents and email threads.

Every deal your team reviews adds to a private knowledge layer that connects deals, portfolio companies, funds, LPs, documents, decisions and source evidence. Analyst corrections and overrides are kept, so a metric definition fixed once stays fixed. Over time Clarenza learns how your firm defines, weighs and decides, and applies that consistently. This memory is tenant-isolated: it belongs to your firm, it is never shared or used to train models, and it is accessible to your whole team.

Yes. Agent reasoning, analyst edits and overrides are logged per deal, together with approval routing and recorded decisions. A deal cannot move to IC until every open condition is answered or accepted. Exportable AI audit reports show what was known, verified and decided at each point, so the decision chain from source to figure to vote can be shown to an IC, an LP or a regulator.

No. Clarenza takes on the programmable, mechanical part of diligence: reading everything, extracting and reconciling figures, running the calculations, prefilling questionnaires and writing the first draft. Your team keeps the judgment: what matters, what to probe, what to believe and what to decide. Analysts stay in control of every output, and because every figure is one click from its source, juniors learn the nuances of a deal faster rather than losing them.

Yes. All analysis of deal data runs on EU-sovereign models inside EU infrastructure. Your deal data never reaches a third-party US model provider and is never used to train models. Clarenza enforces zero data retention on inference, end-to-end encryption (AES-256 at rest, TLS 1.3 in transit), strict tenant isolation and EU-only data residency. Because data stays under EU jurisdiction, it sits outside the reach of the US CLOUD Act and FISA 702. Learn more.

Three. Managed cloud, hosted in the EU, for teams that want to start quickly. A dedicated private space for firms that need isolated infrastructure. And on-premises deployment for firms with the strictest security requirements. All three run the same product. Learn more.

Clarenza is designed for EU AI Act compliance and DORA readiness from the ground up. Full agent reasoning is recorded for every analysis, outputs are traceable to their sources, and human oversight is built into the workflow. This gives your compliance team the documentation it needs instead of asking it to build AI policy, oversight records and a DPIA around a general-purpose tool. Learn more.

SSO and MFA for every user, role-based permissions and deal-level access control, full access logs and instant offboarding. Only people your firm has added can see your deals, and only the deals they are permitted to see.

Yes. Clarenza offers a growing ecosystem of 50+ connectors: data room providers and cloud drives for ingestion, email and meeting capture for deal context, market data such as PitchBook and Crunchbase for enrichment, and CRMs such as Affinity so that findings sync back to your system of record. If a tool you rely on is missing, tell us during onboarding.

You can get started today and create an account here.

First memo on the same day. Configuring your fund thesis takes under an hour with our team, and the first data room can be processed immediately after. Connecting your CRM and other tools can follow at your own pace.

Pricing depends on team size and deployment option. Pick a plan that works best for your fund during signup here.