Skip to content

AI systems built for daily operations

Venice Partners designs, builds and runs AI systems inside client environments, driving operational change and measurable results.

  • Salesforce data
  • AI implementation
  • Document processing
  • Workflow automation
  • Research systems
  • AI governance

Selected work

Systems and programmes delivered by the founders since 2024. Hold or drag a column to read, or filter by sector.

  • Entry 1, Wealth management, Build: Documents, emails and meeting notes reconciled against the existing Salesforce record, with proposed field and related-record changes reviewed before application.
  • Entry 2, Special situations, Run: A private AI platform live inside the fund's own Azure tenant, fourteen services behind the firm's login.
  • Entry 3, Special situations, Build: 17,500 podcast episodes transcribed and made searchable for the investment team, with a weekly digest.
  • Entry 4, Private equity, Train: Thirty team workshops and twenty-five one-to-ones in five months, from the founding partner to fund services.
  • Entry 5, Insurance, Build: Medical underwriting packs of up to two hundred pages summarised for underwriters, in production since 2024.
  • Entry 6, Private equity, Govern: A standing seat on the firm's weekly AI task force: approvals, rejections, security reviews, spend caps.
  • Entry 7, Wealth management, Build: Pension transfer suitability reports drafted from the firm's own templates. Advisers review, edit and sign.
  • Entry 8, Special situations, Build: A research scheduler that runs the team's recurring questions and emails them what changed.
  • Entry 9, Private equity, Build: ChatGPT and Claude connected to SharePoint and Outlook under EU data residency, security-reviewed first.
  • Entry 10, Wealth management, Build: New business compliance checks automated, with a weekly regulatory horizon scan for the compliance team.
  • Entry 11, Special situations, Build: Internal research notes embedded and made queryable from Claude and ChatGPT, inside the firm's permissions.
  • Entry 12, Insurance, Build: Quote and medical evidence vendors integrated straight into Salesforce, retiring a manual back-office process.
  • Entry 13, Private equity, Build: An HR policy agent for staff questions, and RFP drafting for the investor relations team.
  • Entry 14, Wealth management, Build: Provider calls and client meetings transcribed, summarised and filed into the CRM.
  • Entry 15, Special situations, Build: Sector dashboards for chemical trade flows, US property and airline traffic, each with AI chat over the data.
  • Entry 16, Private equity, Train: A firm AI handbook, an internal AI hub, and a five-module training pack for portfolio companies.
  • Entry 17, Wealth management, Build: Pension plan details extracted from provider documents for the transfer team.
  • Entry 18, Insurance, Build: A proofreading assistant for adviser letters and client emails, tuned to the firm's house style.
  • Entry 19, Private equity, Govern: AI spend mapped and managed: licence audits, usage dashboards and a cost controls paper.
  • Entry 20, Wealth management, Govern: A firm-wide AI usage policy and guardrails, then licence guidance for the Copilot rollout.
  • Entry 21, Special situations, Train: Drop-in AI clinics and one-to-one builds with analysts, from NDA review to cash reconciliation.
  • Entry 22, Insurance, Train: An on-site AI maturity assessment with the underwriting back office that set the firm's adoption plan.
  • Entry 23, Private equity, Govern: A firm line on AI and people data: anonymise first, no authored judgements, humans decide ratings.
  • Entry 24, Wealth management, Build: Group risk quotes turned into client-ready reports by a set of template population apps.
  • Entry 25, Insurance, Build: Adviser calls transcribed with notes logged automatically to Salesforce.
  • Entry 26, Family office, Train: A back office enabled on Claude for Excel trade processing.
  • Entry 27, Family office, Train: Workshops for a founding partner's family office, with ongoing support agreed.
  • Entry 28, Wealth management, Govern: A client letter automation scoped twice and never built: the risk outweighed the saving.

The founders

Jack Elton

Co-founder, CEO

Jack has spent fifteen years building and commercialising technology businesses, from early-stage software companies to a global role at JLL.

Read more about Jack Elton

At JLL, he built the Global Business Development function for the firm's largest business line, generating more than US$2 billion of annual qualified pipeline. He previously built Proda's Northern European business, selling an AI platform that converts unstructured property and portfolio data into structured information for banks and asset managers, and worked as a software investor at DN Capital and Frontline Ventures.

At Venice Partners, Jack leads commercial strategy, customer development and the assessment of where AI can produce a clear business return.

John Cruz

Co-founder, CTO

John has spent the past two years designing and operating AI systems in regulated businesses, including Salesforce data completion, document processing, workflow automation, research and controlled generative AI.

Read more about John Cruz

He began his career at Salesforce, where he spent three years as a solutions engineer in enterprise financial services. He left to build directly with generative AI and has since taken systems from initial requirement through production use.

At Venice Partners, John leads product and engineering, with particular responsibility for Salesforce, data, security, human review and production reliability.

How we operate

The full context

We consider the existing system, the underlying information and the relevant business rules together, and aim for a correct operational result.

Evidence before action

Material changes are supported by the information that justifies them, so your team can understand what is proposed before deciding what proceeds.

Within your environment

Systems are designed around your existing software, permissions, data boundaries and integrations.

Accountable in production

We remain involved after launch, monitor performance and improve what is live against actual use.

Where is information being lost between your people and your systems?

We work with organisations where incomplete Salesforce data or manual information-heavy processes are constraining operations, reporting or scale. The starting point is a specific problem and a clear view of the value of solving that problem.