InviscidInviscid — home
Smooth current lines moving across the surface of a dark lake at dusk

Accepting new engagements for Q3

Institutional data infrastructure for lean investment teams.

Inviscid designs, builds, and supports custom data infrastructure and portfolio management systems for hedge funds and family offices — from custodian and broker feeds through to position-level reporting.

Built to your book, not a template
CustomBuilt to your book, not a template
Deployed under your access controls
Your cloudDeployed under your access controls
Ongoing support once you are live
SupportedOngoing support once you are live
Principals only, no juniors staffed on
One teamPrincipals only, no juniors staffed on

Where the friction is

Most small funds don’t have a data problem. They have a plumbing problem.

Every manual hand-off between systems adds drag: another reconciliation, another stale number, another hour that should have gone to research.

  • 01

    Spreadsheets became the system of record

    Positions live in one workbook, exposures in another, and the reconciliation happens in someone’s head at 11pm.

  • 02

    Vendor platforms priced for billion-dollar shops

    You pay enterprise licensing for 20% of the functionality, and still export to Excel to answer a simple question.

  • 03

    No single source of position-level truth

    Prime broker, custodian, and admin all disagree. Nobody can say which number the IC should be looking at.

Data infrastructure

One pipeline, every source, reconciled nightly.

We stand up an ingestion layer that pulls from custodians, prime brokers, fund admins, market data vendors, and your own spreadsheets — normalises it into a single security and entity master, and reconciles it on a schedule you can audit.

  • Custodian, PB, and admin file ingestion with automated breaks handling
  • Security, entity, and counterparty masters with full history
  • Point-in-time warehouse: any number, as it looked on any date
Explore the data platform
Parallel current streaks running across a slow river at dawn

Portfolio management

A portfolio system shaped to how you actually invest.

Public equities, credit, derivatives, funds, direct deals, real assets — modelled properly instead of forced into a template. Exposure, attribution, and risk views are generated from the same book of record your accounting runs on.

  • Multi-asset, multi-entity, multi-currency book of record
  • Look-through exposure for funds, SPVs, and co-invests
  • Performance and attribution down to strategy, sleeve, or holding
See the portfolio engine
A slow eddy circling behind a mossy stone in a forest river

The flow

From raw feed to reportable truth, without a manual step in the middle.

  1. 1

    Ingest

    Custodian and PB files, admin NAV packs, market data, manager statements, your own sheets.

  2. 2

    Normalise

    Mapped to one security, entity, and counterparty master with full effective-dated history.

  3. 3

    Reconcile

    Automated position, cash, and transaction matching with an exception queue people can work.

  4. 4

    Value

    Pricing hierarchy, FX, accruals, and look-through applied to produce the book of record.

  5. 5

    Distribute

    Dashboards, IC packs, LP statements, and an API your other tools can read from.

How we work

A small senior team, embedded alongside yours.

No offshore delivery pods, no year-long discovery phase. The people who design your architecture are the people who write the code, and who you deal with afterwards.

We are a partner, not a vendor.

Feeds change, instruments change, mandates change. We stay available to keep the system correct as they do.

  1. Weeks 1–2

    Diagnostic

    We map every system, file, and spreadsheet in your investment operations, then hand you a written architecture and a build plan with fixed scope.

  2. Weeks 3–10

    Build

    Pipelines, book of record, and reporting shipped in two-week increments. You see working software with your own data from the second increment onward.

  3. Ongoing

    Support

    We monitor the pipelines, absorb custodian and broker format changes, and extend the system as your book evolves.

Let’s talk about where your data gets stuck.

A 30-minute call is usually enough to map your current stack, name the bottlenecks, and tell you whether we are the right build partner.