A PropTech-focused venture firm: Attio CRM build: sourcing, deals, LP & fundraising system

In one line. A PropTech-focused venture firm moved off Affinity onto Attio; we designed the data model and built its sourcing, deal, LP and fundraising systems, landing a 602-record LP object with commitments decoupled from the LP records.

Client & context

The firm invests in construction, infrastructure and real-estate technology, legacy industries its thesis reads as underinvested. The team splits between an investor-relations and operations lead (LP-focused) and the investment side. After migrating off Affinity, the Attio workspace needed a real data model and workflows.

The problem

The workspace needed a clear architecture. One open question: should the deal pipeline live on the company object or in lists? With ~20K company records and only 10-20% of them actual deals, deal attributes on every company would leave much of the workspace empty. Sourcing was manual, the firm tracks 1,500-2,000 companies, and a 313-company tracking bucket needed hygiene work. The LP object mixed people and companies on one list, with no way to track commitments across funds.

What we built

• A VC operations handbook structured around four core functions: deals, sourcing, portfolio management, LPs/fundraising
• An Attio data model: deal object architecture, company as system of record, LP object, lists-based workflow attributes
• A Commitments object decoupled from LP records, the fundraise pipeline runs like a deal board
• A fundraising pipeline for the new fund (602 outreach targets) with commitment and stage tracking
• Spectre/Spectra enrichment on company objects
• Workflow automations (e.g. a 7-day no-interaction Slack notification) and a two-tier pass-email review (comeback vs. not-a-fit)
• A PitchBook CSV → company-record Zapier updater, fed by an upload form
• 2-minute training and change-management videos

How we did it

We worked on a technical-audit-plus-monthly-retainer model with weekly standups. The core recommendation: keep workflow attributes on lists rather than objects, a microservices-vs-monolith argument that keeps company data clean and supports Spectre enrichment. With no PitchBook API access, the upload form feeds a Zapier loop mapping rows onto company records. LP management moved to commitment-based tracking, each LP's fund-by-fund history modelled as separate commitments, resolving the mixed people/companies problem.

Outcome

The commitment decoupling is implemented, new-fund commitments are visible against LP records, and the 602-record LP object is organised (183 individuals) with a master distribution list. The data model, sourcing and deal systems, enrichment and workflow automations are in place on a clean list-based architecture that keeps company data separate from pipeline workflow.

Takeaway: LP object organised (602 records, 183 individuals) with commitments decoupled into a fundraise pipeline, on a clean list-based Attio architecture.