Zalos

What Zalos does

Zalos (YC F25) is the automation layer for the Office of the CFO. Our computer agents transform repetitive, human-driven CFO processes into automated workflows. Tasks such as logging into existing ERPs & portals, handling 2FA & CAPTCHA, automating reconciliations, categorization, and form-filling, and recording every step for audit & compliance are automated. Zalos allows finance teams to save thousands of human hours every year and make more informed decisions with up-to-date data. It builds AI workforces that run finance workflows across fragmented systems, mapping processes, automating them, and continuously improving. Zalos creates a living model of finance operations, making work visible, redesigning finance processes, and generating process guides for SOX compliance and onboarding.

Key capabilities

• AI computer agents for finance workflows
• Automated login to ERPs and portals
• 2FA and CAPTCHA handling
• Automated reconciliations
• Automated categorization
• Automated form-filling
• Process mapping and improvement recommendations
• Self-learning agents that improve with exceptions
• Complete source-to-output lineage recording
• Enterprise-grade security and controls
• Data region and retention policy selection
• No model training on customer data
• Encryption and tenant isolation
• Token-masking on sensitive fields
• Option to run agents entirely inside own environment
• Model agnostic task routing
• Session-level screen recording
• Enterprise-grade audit logs
• Documented, versioned, and cleanly handed over agents
• Incident response plan with 24-hour breach notification for enterprise customers
• Quarterly pen tests and monthly vulnerability scans

Use cases

• Automating month-end reconciliations
• Data ingestion from bank and card statements
• Matching bank transactions to ERP entries
• Flagging unmatched items for accounting teams
• Generating process guides for SOX compliance
• Finance team onboarding
• Understanding how teams run finance processes
• Diagnosing current process issues
• Identifying process improvements
• Adding Controls Agents to improve process performance

AI approach

Zalos builds AI workforces (computer agents) that automate finance workflows across fragmented systems. These agents log into existing ERPs and portals, handle 2FA and CAPTCHA, automate reconciliations, categorization, and form-filling. The agents learn from exceptions and continuously improve. Zalos uses enterprise API licenses from OpenAI, Gemini, and Anthropic, routing tasks to the best-performing model without using customer data for model training.

Tech named: OpenAI, Gemini, Anthropic

What it says sets it apart

• Automation layer for the Office of the CFO
• Computer agents transform repetitive, human-driven CFO processes
• No integrations to build, no rip-and-replace required
• Agents deploy inside existing systems
• Self-learning agents continuously strengthen controls
• Captures complete source-to-output lineage recordings for auditability
• Creates a living model of finance operations
• Converts tribal knowledge into process mappings
• Diagnoses current process issues and recommends improvements
• Generates rigorous process documentation automatically
• Enterprise-grade security and controls with data region and retention policy options
• Customer data is never used to train or fine-tune any model
• Model agnostic approach, routing tasks to the best-performing model
• Prevents 'shadow agent sprawl' with documented and versioned agents
• Only AI finance platform with session-level screen recording alongside enterprise-grade audit logs

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.