Jake Reichert

About Jake Reichert

Concentrated leadership and development experience in both ML/AI application development (data analytics & prediction, vector data storage & prompt augmentation, IAM / SSO) Designing model training & inference infrastructure that minimize GPU costs, expand capacity options, lower training times, and increase durability Develop and roll out AI dev tools across an org in a systematic, measurable way to dramatically improve the execution speed and quality of delivered code Track record of driving transformational organizational change to meet product delivery goals 0-to-1 product development from ideation through deployment at multiple companies Architecting systems to scale to immediate projections Extensive experience with uptime, compliance and quality for very large Enterprise SaaS customers (Hulu, Zoom, Disney, Airbus, Farmers Insurance) Scaled teams from 3 to 80+, moving companies from Series A through Series D

Experience

1. VP of Engineering

   Amesa

   Oct 2024 — Present · San Francisco, CA, US

   AMESA turns customer data into autonomous AI agents that make million-dollar decisions. Using our low-code platform, customers teach expert skills to a team of agents using deep reinforcement learning, traditional ML, and LLMs to build full multi-agent systems.Expanded compute capacity by ~7x and reduced reinforcement learning infrastructure costs by 46% by architecting for hardware-agnostic execution across GPU vendors, spot-tolerant workloads, and intelligent GPU/CPU separation — converting single-vendor shortage into a cost and procurement advantageCut developer hours/week by 57% and ensured every new training capability reached all customers immediately — whether they train on our infrastructure or their own — by consolidating multiple training systems into a single pluggable job execution framework following MLOps practicesImproved engineering time-to-production by 3.17× by evaluating dozens of Large Language Models (LLMs) for coding assistance, distilling them to five with known strengths and cost profiles, and encoding those decisions into developers\' tools — reducing per-engineer AI tooling costs by 34%Personally intervened to save a must-win Fortune 50 deal after an 11th-hour requirement change in under 24 hours by designing, building, and shipping an air-gapped deployment feature, including direct communication with Fortune 50 executive stakeholdersEliminated emerging Fortune 50 deal-blockers by identifying architectural risks in Supabase and directing migration to Postgres and a platform-agnostic architectureIdentified the need for SOC 2 and delivered it from scratch in 3.5 months — initiated 4 weeks before the first customer required it, removing a hard blocker from the Fortune 50 go-to-market pipeline and establishing responsible AI deployment standards for enterprise customers

Education

• University of San Francisco

  BS, Mathematics, Physics

Skills

• Engineering Management
• Agile Methodologies
• Software Architecture
• Java
• Spring Framework
• Spring Security
• Ruby on Rails
• Saas
• Apis
• Analytics
• Javascript
• Jquery
• Mysql
• Redis
• Oracle
• Linux
• Tomcat
• Apache
• Nginx
• Amazon Web Services (Aws)
• Information Architecture
• E-Commerce
• Online Advertising
• Web Applications
• Mobile Applications
• Perl
• Hipaa
• Software Engineering
• Node.js
• Mongodb
• Payment Card Industry Data Security Standard (Pci Dss)
• Docker
• Kubernetes
• Rabbitmq
• Mongo