Job Post: Senior AI Software Engineer
Frank Overton
Job Title: Senior AI Software Engineer
Location: Boulder, Colorado (Remote Considered)
About this Role: FasCat is transforming endurance coaching and training with AI. Our flagship agent, CoachCat, emulates the depth of knowledge and personalized approach of a human coach from a proprietary data set. CoachCat uses sophisticated AI to analyze incoming daily ride and wearable data to dynamically revise training plans in real time, and deliver individualized workout regimens just like a real cycling coach. Our mission is to make coaching more accessible and engaging through AI-driven personalization and actionable insight.
Position Overview: We’re seeking a seasoned AI Software Engineer with expertise in developing and scaling high-impact, real-time AI applications for consumer products. You will help build a platform that helps athletes and coaches accomplish their objectives in a fun, flexible and effective way. This role involves expanding CoachCat’s capabilities by leveraging advanced LLMs, proprietary training data, and reinforcement learning. Your work will enhance multimodal AI interactions to deliver unparalleled insights and responsive, personalized training adjustments.
Responsibilities:
- Architect, design, and deploy advanced AI solutions to power CoachCat, with an emphasis on real-time, data-driven personalization and continuous model learning.
- Develop and optimize ML algorithms for customized training plans that autonomously adapt to athlete performance, powered by data ingestion and continuous model fine-tuning.
- Partner with our engineering team to enhance CoachCat's infrastructure, leveraging transformer models and foundation models to boost response accuracy, interpret athlete-specific metrics, and enable continuous feedback loops.
- Utilize advanced LLM’s, proprietary data, and multimodal models to create a truly interactive and immersive user experience that evolves with each user interaction.
- Maintain a fast-paced, iterative development process, making real-time decisions to deploy updates seamlessly and maintain CoachCat's position as a cutting-edge coaching solution.
Qualifications:
- 4-7+ years of experience in AI and software engineering, specifically with advanced language models, multimodal AI, or reinforcement learning in consumer products.
- Desire to join an exceptionally strong team and accelerate developmental velocity
- Proficiency in machine learning frameworks (e.g., TensorFlow, PyTorch), experience with transformer-based models, and deploying LLMs in production.
- Deep understanding of RLHF (Reinforcement Learning from Human Feedback) and fine-tuning methodologies to create robust, user-aligned AI applications.
- Proficiency in managing data labeling pipelines, including using annotation tools (e.g., Labelbox, Prodigy, Amazon SageMaker Ground Truth) and ensuring high-quality, domain-specific labeled data. Experience with organizing and managing labeled datasets for training, validating, and testing models, especially in applications where precise feedback and accuracy are essential.
- Expertise in designing, refining, and testing prompts for optimal responses from language models. Skilled in prompt engineering techniques to maximize model relevance and context alignment.
- Experience deploying large language models and multimodal frameworks in cloud environments.
- Ability to work autonomously in a fast-paced environment and collaborate effectively with a high-performing engineering team.
- Familiarity with fitness and sports science data or experience in endurance sports is a plus.
Why FasCat Coaching?
- Help Build AI first software to help athletes and coaches be smarter and more productive
- Work on a groundbreaking AI product, creating an intelligent and responsive digital coach that outpaces traditional coaching.
- Collaborate in a mission-driven, AI-first environment focused on reshaping fitness and athlete performance with bleeding-edge technology.
- Flexibility in work environment, with a remote-friendly culture and a strong focus on innovation and team success.
Contact
Interested Applicants with experience and background that meet or exceed all the criteria above are encouraged to reach out to the CEO, Frank Overton frankatfascatcochingdotcom.
Please include a description of your most recent work, background and resume.
About FasCat Coaching
Headquartered in Boulder, Colorado, FasCat Coaching is committed to delivering high-quality coaching to athletes at all levels, empowering them to have fun and achieve peak performance. With a dual approach—AI Coaching and 1x1 Coaching—FasCat blends cutting-edge technology with professional expertise. The company’s CoachCat AI draws on over two decades of experience and a proprietary data set to provide personalized insights, while FasCat's team of elite coaches brings firsthand racing experience, data-driven strategies, and a dedication to clear, effective communication
About CoachCat
CoachCat is the first hardware-agnostic AI coaching app tailored for cyclists, to deliver a complete, personalized training experience. This innovative app customizes training plans by seamlessly integrating workout data from cycling computers and wearables—such as Apple Watch,Wahoo, Oura Ring, Whoop, Garmin, and more—to provide athletes with a 360-degree view of their physiological strain and recovery. With instant, real-time coaching feedback, adaptive workouts, and dynamic training plan adjustments, CoachCat empowers cyclists to optimize every ride and make informed progress toward their performance goals.
Wintertime Workouts: the Cheetah Pounce & Sprint > Stalk > Catch ‼️
Recovery for Cycling
Road Racing Intervals
- Increase your functional and race-specific power output
- Includes Sweet Spot, VO2, Anaerobic, Threshold
Road Race In-Season
- Weekend racing and group rides with weekday training and recovery
- anaerobic efforts like criss cross, Over/Unders Sweet Spot, Threshold
Foundation : 3 Weeks
- Perfect for all cyclists beginning off season training
- Raise your CTL and the all-important muscle tension intervals
Phil Gaimon's Strava PR Plan
- Perfect Plan for Those with Less Training Time, starts at 15 minutes per day
- VO2's, 1 minuters, Tabatas, threshold, suprathreshold, and even Sweet Spot