Pipeline fragmented across a dozen systems
CRM, Outlook email, OneDrive, shared spreadsheets. The same founder lives in 4 different records. Figuring out 'where are we with X' takes 20 minutes of context-switching.
Finance Ops · Venture Capital
Deal flow run by AI agents.
You focus on deals.
The agent monitors emails, archives with metadata, deduplicates deals, multi-criteria scoring, auto-generated monthly report. Compliant, zero operations overhead. Live in 4 weeks.
If it doesn't work, you don't pay. Guarantee included.
100% emails archived
0 hours of manual deduplication
Monthly report auto-generated
Soraia · agent activity ·
·
·
Already chosen by Italian funds
LIFTT · Deep tech Venture Capital holding · active client
Are you a non-VC finance team (accountants, CFO, M&A operations)? Go to the Finance Ops page
The problem
Deal flow has grown faster than the processes that manage it.
A dozen systems, a thousand email threads, countless files on OneDrive. The senior team spends more time hunting for data than doing deals.
Emails not archived for due diligence
When you need to reconstruct the history of a deal, the key conversations are buried in some Outlook thread nobody can find. Compliance issue.
Monthly report built by hand
Every start of the month, hours of qualified work to produce pipeline numbers, conversion rates, activity summaries. By the time the report is ready, it's already stale.
The solution
6 AI agents that form the backbone of your deal flow.
Monitoring, archiving, deduplication, scoring, reporting, portfolio tracking. End-to-end digital workers. The team steps in only for decisions that require judgement.
Deal-related email monitoring
The agent reads the deal team's mailboxes and automatically recognises what is deal-related versus spam, newsletters, and internal communications.
Compliant email + attachment archiving
Eml + pitch deck + term sheet archived on OneDrive/SharePoint with structured metadata (deal name, founder, date, attachment type). Ready for due diligence.
Cross-source deal deduplication
The agent recognises that "Mario Rossi - Startup A" and "M. Rossi (CEO StartupA)" are the same record. It proposes the merge; a human approves with one click.
Multi-criteria deal scoring
For every new deal in the pipeline, the agent cross-references Crunchbase, PitchBook, and founders' LinkedIn profiles, and produces a justified score for the deal team.
Auto-generated monthly report
Auto-updated charts at the start of each month: pipeline by stage, conversion rate, activity by deal team member, alerts on stagnant deals. No more manual Excel exports.
Portfolio monitoring + alerts
For portfolio companies, the agent tracks KPIs reported by founders, alerts on deviations, and sets calendar reminders for periodic check-ins. Nothing slips through the cracks.
Our guarantee
Autonomous deal flow backbone.
Measured in 4 weeks. Or you don't pay.
Baseline measured during the assessment (hours/person/week on monitoring + deduplication + reporting). If the agent does not recover the promised hours, we work for free until it does — or we refund the sprint.
Case study
AI agents in action at LIFTT.
Comparison
Why custom AI agents, and not something else.
The three real options for a VC fund or family office today.
| ChatGPT / Copilot | Affinity / Attio AI | Soraia AI agents | |
|---|---|---|---|
| Cost | Low | High annual | Medium one-off |
| Time to go live | Immediate | 3–6 months | 4 weeks |
| Executes tasks autonomously | No | Limited | Yes, end-to-end |
| Custom to your deal flow | No | Configurable | Built for you |
| Vendor dependency | None | Total | None (your code) |
| Email archive compliance | No | Partial | Yes, audit-ready |
Questions
What people usually ask us.
How do you handle GDPR and confidential deal data?
European or on-premise infrastructure on your own stack. No training on your data. Full logs, configurable retention. Already in use on deal flows of regulated Italian funds.
Does it integrate with Affinity, Salesforce, Attio, or a custom CRM?
Yes. We work via API or programmatic export. If you have a custom CRM or Airtable, we assess it during the assessment phase.
How does email archiving for due diligence work?
Eml + attachments on OneDrive/SharePoint with structured metadata (deal, founder, date, type). When you need to reconstruct a deal's history, two clicks find everything. Compliance-ready.
What about portfolio monitoring?
The agent tracks the KPIs founders send you (monthly, quarterly), structures them, alerts on deviations, and sets calendar reminders for check-ins. Zero KPIs lost in your inbox.
What does it actually cost?
Assessment €2,000 (refunded if you proceed). Sprint €10–50k. Guarantee: if the agent does not recover the hours promised to the deal team, we work for free until it does.
Next step
Find out how many hours your deal team can get back.
Assessment €2,000, refunded if you proceed. 1–2 weeks from kick-off to report.
