Software for funds that
outgrew spreadsheets.

An engineering shop. The work is software — pipelines, automations, internal tools, infrastructure — not trade ideas, portfolio construction, or risk analysis.

What's unusual is that the engineering comes with domain context built in: fifteen-plus years of production software, plus a working systematic trading infrastructure behind the practice. That means no three-month ramp to understand the deck, and no guessing at which problems are worth solving in software.

Most engineering consultants have to learn finance on a client's dime. This one doesn't.

A Spike is a discrete thing: a reconciliation pipeline, a thesis exposure tracker, a reporting automation, a specific framework encoded into software. One clear deliverable, four to six weeks, fixed price.

A Retainer makes sense when you have a series of smaller projects or ongoing integration work — pipeline maintenance, new broker onboarding, incremental features, monthly ops tooling requests. Five to ten hours a week, month-to-month.

Most funds start with an Audit, then a Spike. Retainers usually come after we've already shipped something together.

Quant devs build models and execution infrastructure for quant funds. That's a specialist skillset with specialist rates and specialist priorities.

Discretionary funds usually don't need a quant dev. They need someone who can wire Bloomberg to Addepar, automate the reconciliation spreadsheet, and turn the PM's investment framework into software the ops team uses. That's engineering, not research.

Ansatz is the second kind. If a project genuinely needs a quant dev, we'll say so.

Discretionary funds — macro, commodities, event-driven, long/short equity — in the roughly $200M–$2B range, with two common starting points: operations that have outgrown the spreadsheet stack, and an investment framework that lives in the PM's head instead of in software the team can actually use.

The work takes two shapes, usually in combination. Ops Automation: reconciliation, pipelines, reporting, broker integrations, internal tools. Investment Process in Software: thesis tracking, exposure monitoring, analyst workflows — whatever the discipline requires.

Quant funds are rarely the fit for core quant work, but non-quant infrastructure — investor reporting, data pipelines, internal tooling — is fair game. If you're not sure, raise it on the call.

NDAs are standard and expected. Ansatz signs yours.

Access is scoped to what the project needs. For an ops audit, read-only access to representative systems is usually sufficient. For a Spike, we'll define access narrowly — specific broker credentials, specific database views, specific folders — rather than blanket admin.

Data handling and privacy practices are available on request.

That's the default state. That's why we're talking.

Part of the Audit is exactly this assessment: what the data actually looks like, what's clean enough to build on, what needs to be normalized first. A lot of the initial value of a Spike Project is often just imposing structure on the mess.

If the data is so chaotic that the first engagement has to be a normalization pass, we'll say so up front and scope accordingly.

Python for data work, TypeScript and modern web stacks for tooling and internal apps, SQL everywhere, whatever cloud infrastructure your fund is already using.

Existing vendor systems — Bloomberg, Addepar, Enfusion, prime broker portals — stay in place. The value is usually in the connective tissue between them, not in a rebuild.

If you're on a legacy mainframe system or something deeply specialized, we'll talk about fit before we scope.

No.

The Audit is a one-time fixed fee. Spike Projects are fixed-scope with clear deliverables and clear end dates. Retainers are month-to-month with thirty-day cancellation.

The goal is to be useful, not resident.