The archive · Money & Fintech · Product decision · 2024-2026
Stacks' agentic-close bet: AI agents reconcile the books; Lightspeed leads $23M
Ex-Uber and Plaid product leader Albert Malikov's Stacks automates enterprise month-end close on an AI-ready data layer; Lightspeed led a $23M Series A.
Stacks
What the business is
Stacks builds AI agents for enterprise finance: a data layer connects to every financial system an organization runs, giving one consistent view, while agents automate month-end close, account reconciliations, journal entries, variance analysis and reporting in sync with the ERP.
Starting capital：$12M seed (early 2025), then a $23M Series A led by Lightspeed with General Catalyst, EQT Ventures and S16VC (February 2026), bringing total funding to $35M
How it started
Albert Malikov founded Stacks in 2024, initially in Amsterdam and now headquartered in London, after product leadership roles at Uber and Plaid, where he scaled Plaid's European business. His thesis was that enterprise financial operations were crippled by data scattered across ERPs, spreadsheets, data lakes and legacy systems, forcing finance teams into manual workarounds because core platforms were not built for AI.
What happened
Stacks closed a $12M seed round in early 2025 and built a data layer connecting directly to finance systems, then deployed agents with deterministic reasoning to automate monthly close operations, reconciliations, journal entries and variance analysis. By February 2026 it had more than 30 enterprise customers and about 100,000 hours saved, and on February 19, 2026 it announced a $23M Series A led by Lightspeed with participation from General Catalyst, EQT Ventures and S16VC, alongside the launch of an AI Flux Analysis suite that replaces spreadsheet-based commentary with explainable, account-level investigation.
How it ended up
With the Series A, Stacks was moving from close automation toward 'financial intelligence', launching AI Flux Analysis for variance analysis and an executive-summary tool, while aiming at the office-of-the-CFO software market it estimates at more than $100 billion a year.
Background
Stacks is a London-headquartered startup that automates enterprise financial operations with AI agents. Its core idea is that fragmented data is the real obstacle: transaction-level detail sits across ERPs, spreadsheets, data lakes and legacy tools, so finance teams burn time on reconciliation instead of analysis and planning.
Founder Albert Malikov, previously a product leader at Uber and Plaid, started Stacks in 2024, building a data layer that connects to every financial system in an organization to create a single, consistent view. On top of that foundation, Stacks deploys agents with deterministic reasoning that automate month-end close, account reconciliations, journal entries and variance analysis, synchronized with the organization's ERP.
Stacks raised a $12M seed round in early 2025 and grew to more than 30 enterprise customers globally, including publisher Future and audio company Epidemic Sound, reporting about 100,000 hours saved. On February 19, 2026 it announced a $23M Series A led by Lightspeed, with General Catalyst, EQT Ventures and S16VC participating, and launched AI Flux Analysis, which automates variance analysis with explainable, account-level investigation.
The bet is that CFO teams will hand repetitive close work to agents and keep people for judgment: 'From day one, we focused on solving the core problem of fragmented data,' Malikov said. Stacks is targeting the office-of-the-CFO software market, which it estimates is worth more than $100 billion a year even before counting the labor spent keeping accounts straight.
What has to be true
- Stacks chose the most manual and expensive finance workflows, accounting and month-end close, where automation savings are easy to measure in hours and error reduction.
- It built the data integration layer first, then added agents, an architecture that fixes fragmentation instead of bolting AI onto broken workflows.
- The founder's background at Uber and Plaid gave the team credibility with enterprise buyers and investors, and the startup showed real traction with 30-plus customers before raising its Series A.
- Deterministic reasoning agents plus explainable variance analysis answer the compliance question that keeps CFO teams from trusting black-box AI.
What can be applied
Make the data problem the product: an AI-ready layer that unifies fragmented financial data turns a scattered mess into automation competitors cannot easily replicate.
Aftermath
After the Series A, Stacks planned to move from close automation into broader financial intelligence, with AI Flux Analysis for variance analysis and a tool that generates executive summaries of financial operations. Its agents remained focused on reconciling transactions, journal entries and reporting while staying synchronized with enterprise ERPs, and the funding was earmarked for expanding the platform and go-to-market.
Sources
- Agentic finance automation startup Stacks raises $23M in funding
- Lightspeed leads $23M investment in AI startup for accounting departments Stacks
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card