Swifty

Converts unstructured natural language descriptions of business expenses (e.g., 'lunch with client at steakhouse, $45') into structured expense records with automatic category assignment, amount extraction, and merchant identification. Uses NLP entity recognition to parse dates, amounts, and merchant names from conversational input, then maps to predefined corporate expense categories (meals, transport, accommodation, etc.) without requiring manual form filling.

Aggregates flight, hotel, and meeting information from multiple sources (email, calendar, booking confirmations) into a unified itinerary view accessible via chat. Monitors for schedule changes, delays, or conflicts and proactively alerts users through the chat interface. Uses calendar integration and email parsing to extract travel details and cross-reference with booking systems to detect discrepancies or overlaps.

Validates expenses and travel decisions against company-defined policies (e.g., maximum meal spend per day, approved hotel chains, airline preferences) by analyzing submitted expenses and itineraries in real-time. Stores policy rules as configuration and applies them during expense categorization and itinerary review, flagging violations with explanations and suggesting compliant alternatives.

Maintains a persistent context window that aggregates data from multiple sources (email, calendar, previous chat history, expense records, itineraries) to provide coherent responses to travel and expense queries. Uses a context management layer to prioritize recent information, resolve conflicts between sources, and maintain state across multiple chat turns without requiring users to re-provide information.

Generates personalized travel recommendations (hotels, restaurants, transportation options) based on user preferences, past travel patterns, budget constraints, and policy compliance. Uses conversational context and historical data to suggest alternatives when initial choices violate policy or exceed budget, with explanations for why alternatives are recommended.

Allows users to upload or reference receipt images within the chat interface, storing them as attachments linked to expense records. Provides a centralized receipt repository accessible through chat queries, enabling users to retrieve receipts for specific expenses without managing separate file systems or email folders.

Generates automated expense reports and summaries from aggregated expense records, with breakdowns by category, date, and trip. Produces reports in multiple formats (chat summary, downloadable PDF, email-ready format) suitable for reimbursement submission or budget analysis. Uses aggregated expense data to calculate totals, identify spending patterns, and flag anomalies.

Enables multiple team members to share itineraries, expenses, and travel information within a shared Swifty workspace, with role-based access controls (employee, manager, finance). Provides visibility into team travel schedules, aggregate spending, and policy compliance across the group. Uses shared context and data aggregation to coordinate group trips and identify overlapping travel.