Navi AI
Navi AI, a San Francisco start-up founded in 2024, uses generative AI to turn flight data and cockpit audio into structured debriefs for pilot training.
With up to ten staff, it emerged from stealth in March 2026 with a 6.7 million US dollar seed round. Its products include Cadet for debriefs, Bastion for fleet-wide oversight at flight academies, a device that streams cockpit data and ARES for US Air Force training. No customers are named.
No verified deployments yet
Navi AI is profiled in the companies register, but no deployment has been verified against published evidence. Records are added as evidence is reviewed and graded.
Products & Solutions
4 productsCadet
AI-powered pilot assistant for student pilots in flight training.
Manual post-flight debriefing in flight training is time-consuming, inconsistent, and relies on instructor memory rather than full flight data. Cadet automates and standardises debriefing so trainee pilots receive data-driven, personalised feedback after every sortie.
- Automated flight debrief with 40 to 50 key insights per sortie
- Phase-by-phase breakdown of each flight from engine start to shutdown
- Maneuver identification and performance scoring against school SOPs
- Context-aware AI assistant grounded in POH, PHAK, FAA regulations and flight data
2 more features
- Full sortie replay with time-synced audio, maneuver and correction tracking
- Safety metrics including G-load, minimum altitude and airspeed deviations
Technical specifications (5)
- Domain-specific large language model (LLM) trained on more than 100,000 real flight hours
- Multi-model architecture: Gemini, Llama, Whisper (audio transcription), plus proprietary flight data models
- Inputs: cockpit audio, aircraft telemetry (attitude/speed/altitude), engine parameters, ATC communications, GPS track, ADS-B traffic and weather (METARs, TFRs)
- Wave hardware device: wireless aviation vector endpoint for in-cockpit data capture
- Garmin avionics ecosystem integration
Bastion
Flight training command platform for flight academies and training organisations. Provides institution-level visibility into fleet safety trends, instructor performance analytics, student progression tracking, and real-time operational monitoring across all active sorties.
Flight academies lack aggregate data-driven oversight of fleet-wide safety trends, instructor standardisation gaps, and at-risk student identification before incidents occur. Bastion provides always-on safety management and operational intelligence at the programme level.
- Real-time operational command view of all active sorties, instructors and fleet
- Incident hotspot mapping and trend analysis
- Instructor performance analytics including grading drift and coaching inconsistencies
- Student progression intelligence from first flight to checkride
2 more features
- Automated detection of unstabilised approaches, control transfers and deviation patterns
- CFI intervention rate tracking and transfer-of-controls protocol compliance monitoring
Technical specifications (3)
- Integrates cockpit audio analysis with flight data and environmental context
- Continuous hazard identification and risk trending
- Extends traditional FOQA with instructor and environmental data layers
ARES (Aviation Reasoning Engine for Sortie Intelligence)
AI-powered debrief and flight analysis platform purpose-built for the United States Air Force. Delivers sortie-level intelligence for military pilot training to accelerate pilot development, strengthen mission effectiveness and reduce the combat-ready pilot pipeline gap.
The US Air Force faces a combat-ready pilot pipeline gap. Manual post-sortie debriefing is slow, incomplete and insufficient to accelerate pilot development at the pace and precision required at elite test pilot training institutions.
- Sortie-level AI debrief and performance intelligence for military aircraft
- Adapted for high-precision, high-consequence training environments
- Accelerated pilot development through automated analysis
- Mission effectiveness analytics
Technical specifications (3)
- Adapted from commercial Cadet/Bastion architecture for US Air Force use
- Supports Northrop T-38 trainer aircraft
- Funded under $1.27M US Department of War SBIR Phase II contract
- US Department of War SBIR Phase II contract
Wave (Wireless Aviation Vector Endpoint)
Hardware data capture device installed in training aircraft to collect and transmit cockpit audio, aircraft telemetry, engine parameters, ATC communications, traffic and weather data, and GPS track. Enables the Cadet and Bastion software platforms to function by providing the multi-source data stream required for AI analysis.
Standard aircraft do not continuously stream the full range of cockpit and flight data needed for AI-powered post-flight analysis. Wave closes this gap by acting as an onboard sensor and data aggregation endpoint.
- Cockpit audio capture (all crew and ATC communications)
- Aircraft telemetry: attitude, speed, altitude
- Engine parameters: RPM, temperatures, fuel flow
- Live ATC frequency capture
2 more features
- ADS-B traffic and weather (METARs, TFRs)
- 4D GPS flight path recording
Technical specifications (2)
- Wireless transmission of multi-source flight data
- Integrates with Garmin avionics ecosystem
Profile data from the Aerostack companies register; deployment records from the verified register. Evidence grades: T1 first-party confirmation, T2 credible reporting. See Our approach for how records are graded.