---
title: 'ML/AI Founding Engineer at Navi AI'
canonical: 'https://feeny.ai/job/ml-ai-founding-engineer-navi-ai-san-francisco-r77ng00bz1df'
type: 'job'
last_seen: '2026-09-08'
---

# ML/AI Founding Engineer at Navi AI

- **Company:** Navi AI
- **Location:** San Francisco, CA
- **Employment:** full-time
- **Work type:** onsite
- **Posted:** 2025-10-28
- **Last confirmed live:** 2026-09-08
- **Apply:** https://jobs.ashbyhq.com/navi/f04ee2df-2e9d-4137-92d2-52ec725f015f

## Job description

Navi captures everything a pilot sees and hears and turns it into automated debrief intelligence. You're building the AI that makes sense of it all — the models that listen to cockpit audio, identify maneuvers, attribute instructor vs. student behavior, and generate debrief reports that match what a seasoned CFI would catch.

## ABOUT THE ROLE

This is a founding AI/ML role. You'll own the intelligence layer that sits at the core of everything Navi does — from audio diarization and speech recognition to maneuver detection, safety event classification, and automated debrief generation. The data is messy, multimodal, and high-stakes. Cockpit audio is mono with overlapping speakers. Avionics telemetry arrives in dozens of formats. ATC comms bleed into crew conversation. Your job is to make machines understand all of it — and get it right, because the output goes directly to pilots, instructors, and military operators.

Navi is live at Embry-Riddle, Purdue, UND, Sling Pilot Academy, and the United States Air Force. Your models are already in production. This isn't research — it's AI that flies.

## WHAT YOU'LL DO

- Build and improve the ML systems that power Navi's automated flight debrief — maneuver detection, performance scoring, safety event identification
- Develop and refine audio intelligence pipelines — speaker diarization, speech-to-text, cockpit audio separation, ATC communication extraction
- Design the AI reasoning layer that synthesizes avionics data, audio, and ADS-B into coherent sortie narratives
- Build evaluation frameworks and feedback loops that continuously improve model accuracy against real-world CFI assessments
- Work directly with pilots and instructors to ground-truth model outputs and close the gap between what the AI sees and what actually happened in the aircraft
- Push the boundary on what's possible with LLMs in safety-critical, domain-specific applications

## ABOUT YOU

- 5+ years of professional experience in machine learning, AI, or applied research — with production systems, not just papers
- Deep expertise in at least one of: NLP/speech processing, audio ML, time-series analysis, or multimodal reasoning
- Experience building and deploying ML pipelines end to end — data ingestion, model training, evaluation, inference, and monitoring in production
- Strong engineering fundamentals — you can build the infrastructure your models need, not just the models
- You've worked with messy, real-world data and know how to build systems that are robust to noise, edge cases, and domain drift
- Experience with LLMs — fine-tuning, prompt engineering, retrieval-augmented generation, or building LLM-powered applications
- You ship. You don't wait for a perfect dataset or a clean abstraction. You build, evaluate, iterate, and improve
- Comfortable operating with high autonomy in a fast-moving environment where the problem definition evolves weekly

## NICE TO HAVE

- Familiarity with aviation systems, flight training operations, or defense technology environments
- Experience with audio diarization, speaker separation, or cockpit/radio audio processing
- Background in safety-critical ML systems where model accuracy has real-world consequences

## WHY THIS ROLE MATTERS

FOQA sees a perfect steep turn. Navi sees a disaster. The difference is the intelligence layer you're building — the AI that hears the instructor take the controls two seconds before a G-load exceedance, that knows the student was coached through an approach instead of flying it solo, that catches what the numbers alone will never tell you. This is the system the aviation industry has never had. You're building it.

## WHAT YOU'LL GET

- Early-stage equity — real ownership in a category-defining company
- Flight training — earn your pilot's license and build with true domain expertise
- Impact you can see — your work will be used by pilots, flight schools, airlines, and the U.S. Air Force
- A role that scales into technical leadership as we grow

## HOW WE WORK

- Find a way. We don't wait for permission or perfect information. Ideas come from anywhere regardless of title. Figure it out, ship it, iterate.
- Creativity over control. First principles over process. We'd rather have a creative solution that's 80% right today than a perfect one next quarter.
- Update fast. Come in with a hypothesis, throw it away when the data says otherwise. Ego has no place here.
- Intensity with focus. We work hard because the mission demands it. Clarity on what matters is how we make that sustainable.

## About Navi AI

## Company Overview
- **One-liner**: Navi AI builds the first purpose-built generative AI platform for aviation, providing an AI co-pilot for pilot training that turns every aircraft into a data source to deliver automated, data-driven flight debriefs and safety intelligence.
- **Entity Type**: Private (Seed stage)
- **Headquarters**: San Francisco, California, USA (also has an office in Bengaluru, India)
- **Founded**: 2024
- **Founders**: Nikola Kostic (CEO) and Vivek Velivela (CTO)

## Core Business
- **Primary Industry**: Aviation Technology / Aerospace / Defense / Artificial Intelligence
- **Target Customers**: B2B — flight schools (Part 141), universities (aerospace programs), and military/defense organizations (U.S. Air Force Test Pilot School).
- **Mission/Purpose Statement**: To make aviation safety proactive — starting in training. Navi gives every pilot and squadron the intelligence edge they deserve.

## Products & Services
- **Navi AI Platform**: An AI-driven platform that ingests cockpit audio, aircraft telemetry, and environmental data (weather, traffic, aircraft history) to generate moment-by-moment, automated flight debriefs. It provides 40–50 key insights per flight through text, visuals, and animations. The platform is grounded in the flight school’s syllabus, POH, PHAK, and FAA regulations.
- **AI Debrief Assistant**: A context-aware conversational AI that trainee pilots can query about their performance, referencing their own flight data, standard operating procedures, and FAA regulations.
- **Automated Maneuver Analysis**: Analyzes intent, behavior, and performance for every maneuver during a flight, aligning outputs to the specific lesson or maneuver being taught.
- **Full Sortie Replay**: Provides a phase-by-phase breakdown of every training flight from engine start to shutdown.
- **Institutional Analytics (FOQA+)**: Goes beyond traditional FOQA by combining flight data with instructor interactions, environmental data, and procedural context, giving flight academies real-time visibility into trainee progression, trend analysis, and early safety pattern detection.

## Market Standing
- **Valuation**: Not publicly available (seed-stage company).
- **Key Metric – Total Funding**: $6.7 million raised to date.
- **Notable Investors/Partners**:
    - **United Airlines Ventures** (strategic investor)
    - **BVVC** (lead investor)
    - **New Vista Capital**
    - **Raptor Group**
    - **I2BF**
    - **U.S. Department of War** ($1.27M SBIR Phase II grant)
    - **Embry-Riddle Aeronautical University** (strategic partner and equity stakeholder)
    - **Garmin** (technology partnership, integrated with their avionics ecosystem)
- **Growth Signals**:
    - Launched from stealth in March 2026 with $6.7M in funding.
    - First commercial deployment in September 2024 with Sling Pilot Academy; now logging 55,000+ flight hours annually at Sling.
    - Platform deployed or under evaluation at leading flight schools: Embry-Riddle Aeronautical University, University of North Dakota, Purdue University, Utah State University, Delta State University, and the U.S. Air Force Test Pilot School.
    - Plans to hire 15 additional employees in 2026 (currently 7 employees), aggressively scaling the team.
    - Has a $1.27M SBIR Phase II contract from the U.S. Department of War to adapt its platform for the T-38 fleet at Edwards AFB.

## Competitive Advantages
- **First-Mover in AI-Powered Debriefing**: No other company offers a purpose-built, commercially operational generative AI platform specifically for pilot training that automates post-flight analysis.
- **Domain-Specific LLM**: Trained on 100,000+ real flight hours, analyzing intent, behavior, and performance — not just generic AI.
- **Deep Institutional Partnerships**: Already embedded in the world's largest aerospace university (Embry-Riddle) and working with the U.S. Air Force Test Pilot School, providing credibility and a strong go-to-market signal.
- **Strategic Backing**: Backed by United Airlines Ventures (a direct channel to commercial aviation) and the U.S. Department of War, giving it a dual commercial and defense pathway.
- **Real-World Traction**: 55,000+ flight hours annually processed at Sling Pilot Academy, proving the product works at scale in a live training environment.

## Strategic Focus
- **Near-term**: Dominate the U.S. pilot training market by equipping every training aircraft with the Navi platform.
- **Medium-term**: Expand into commercial aviation, applying the same real-time analysis and performance intelligence to airline operations.
- **Defense**: Adapt and scale the platform for military use (U.S. Air Force), starting with the Test Pilot School and the T-38 fleet.
- **Product**: Continue to deepen the integration with Garmin avionics and expand the platform's ability to ingest more data sources.

## Why Work Here
- **High-Impact Mission**: Work directly on making aviation safer, with a tangible product already in use by pilots and the military. The company culture emphasizes building for a critical national need ("Your country needs you").
- **Early-Stage, High-Growth**: With only 7 employees and a plan to hire 15 more, this is a true seed-stage startup where new hires will define the product, culture, and engineering practices from the ground up.
- **Engineering-Centric**: The open roles (Founding Software Engineer, ML/AI Founding Engineer, Founding Hardware Engineer, Deployment Engineer) indicate a deep focus on core product engineering and infrastructure.
- **On-Site / Hybrid**: San Francisco headquarters is primarily in-office; Bengaluru office also exists. The company favors on-site work for its core team.
- **Fast-Paced Culture**: The careers page signals an expectation of high agency: "if ten things a day sounds like a slow Tuesday - we need you."
- **Interesting Tech Stack**: Working with cockpit telemetry, audio, domain-specific LLMs, avionics integration (Garmin), and real-time data pipelines — a mix of hardware, software, and AI.
- **Founding Team Access**: Opportunity to work directly with the co-founders (Nikola Kostic, CEO; Vivek Velivela, CTO) and shape the company's technical direction.

## Sources
1. [flynavi.com](https://www.flynavi.com/)
2. [flynavi.com/careers](https://www.flynavi.com/careers)
3. [builtin.com](https://builtin.com/company/navi-ai)
4. [flynavi.com/newsroom](https://www.flynavi.com/newsroom)
5. [prnewswire.com](https://www.prnewswire.com/news-releases/navi-ai-emerges-from-stealth-to-accelerate-pilot-training-with-ai-302724509.html)
6. [jobs.ashbyhq.com/navi](https://jobs.ashbyhq.com/navi)

## Other roles at Navi AI

- [Intern - Device Build](https://feeny.ai/job/intern-device-build-navi-ai-san-francisco-xscsg8yw7qf1) — San Francisco, CA
- [Part-time CFI](https://feeny.ai/job/part-time-cfi-navi-ai-san-francisco-rg9jj8vsp9qs) — San Francisco, CA
- [Infrastructure Engineer](https://feeny.ai/job/infrastructure-engineer-navi-ai-san-francisco-jc3wsjez8qm6) — San Francisco, CA
- [Founding Go-to-Market Lead](https://feeny.ai/job/founding-go-to-market-lead-navi-ai-san-francisco-dfrx47rngsm4) — San Francisco, CA
- [Business Development Lead](https://feeny.ai/job/business-development-lead-navi-ai-san-francisco-0h8z796s0x6q) — San Francisco, CA
- [Founding Hardware Engineer](https://feeny.ai/job/founding-hardware-engineer-navi-ai-san-francisco-w69bj0taaqav) — San Francisco, CA
- [Forward Deployed Engineer](https://feeny.ai/job/forward-deployed-engineer-navi-ai-san-francisco-px34xzw2j97k) — San Francisco, CA
- [Founding Software Engineer](https://feeny.ai/job/founding-software-engineer-navi-ai-san-francisco-akpvm8bmqk5e) — San Francisco, CA