Location: 100% Remote
Duration: End of 2026, strong possibility of extension
ROLE SPECIFICATIONS
Experience Level: 5 6+ Years in Software & AI/ML Engineering
Primary Objective: Technical Rescue, Agentic Remediation, and Client Turnaround
Role Overview We are recruiting an elite Forward Deployed Engineer (FDE) to join our high-impact AI Seal Team. When a high- stakes customer deployment turns 'RED'-whether due to complex agent failures, unhandled non-deterministic LLM edge cases, or broken client trust-the FDE will drop in, take total technical ownership, and lead the account back to 'GREEN.'
As a Black Belt AI Fixer, the FDE will not advise from the sidelines. The FDE is required to dive deep into production codebases, re-architect failing agentic workflows, resolve non-deterministic LLM behavior, and restore immediate confidence with senior client executives across enterprise domains like Banking, Healthcare, Finance, and Retail.
Key Responsibilities Rapid Account Turnaround ('Red-to-Green')
- Firefight & Fix: The FDE is required to deploy directly into struggling customer accounts, diagnose root causes of technical and delivery friction within 48 72 hours, and execute immediate hands-on remediation plans.
- Lead from the Front: The FDE will take technical command of the existing account team, establish rigorous engineering standards, and rebuild project momentum.
- Executive Client Presence: The FDE will serve as the primary technical contact for client CXOs and lead architects, communicating complex technical trade-offs with clarity, composure, and authority under pressure. AI Agent Engineering & Code Remediation
- Agentic Recovery: The FDE is required to debug and resolve complex failures in production AI agents using Google GECX, G3, and Gemini stack frameworks.
- Deterministic Fallback Architecture: The FDE will design and implement robust, deterministic fallback mechanisms to ensure system reliability when LLMs hallucinate, timeout, or fail guardrails.
- Scope Containment: The FDE will re-architect unconstrained agent prompts into tight, deterministic workflows with strict API/schema validation.
- Hands-on Refactoring: The FDE is required to inspect logs, trace execution paths, and refactor client-side microservices (Python, TypeScript, vector databases) in high-throughput environments.
Required Technical Expertise
- Core Software Engineering (5 6+ Years): The FDE must possess extensive production experience in Python, TypeScript/JavaScript, or Go, backed by strong GCP cloud-native microservices expertise.
- AI & Agentic Systems: The FDE is required to have practical experience building, fine-tuning, and debugging LLM applications, RAG pipelines, and agent frameworks (e.g., LangChain, LlamaIndex, AutoGen, or custom orchestrators).
- LLM Failure Mode Mitigation: The FDE must demonstrate proven capability in building deterministic wrappers around non-deterministic LLM outputs (schema enforcement, tool-use validation, structured JSON parsing, and fallback state machines).
- Enterprise Integration: The FDE must have experience integrating AI agents with complex backend legacy APIs, CRMs, and databases in regulated industries (PCI, HIPAA, Banking compliance).
Mindset & Behavioral Requirements
- Hacker Mentality: The FDE must be pragmatic, highly resourceful, and comfortable navigating messy or undocumented legacy codebases without waiting for detailed specs.
- Battle-Tested Composure: The FDE is required to maintain complete composure in high-stress firefighting scenarios and de-escalate friction with demanding stakeholders.
- Uncompromising Ownership: The FDE must exhibit a relentless 'get-it-done' attitude and take personal accountability for moving the account status from red to green.
- Rapid Adaptability: The FDE is expected to rapidly ingest and master new tech stacks (such as the Google GECX / G3 framework) through intensive hands-on bootcamps.
Preferred Qualifications
- Prior experience operating as an FDE at a top-tier AI company, cloud provider, or elite technical consulting firm.
- Hands-on experience implementing domain-specific AI guardrails for fraud prevention, auditability, or medical compliance.
- Background in real-time observability, telemetry, and evaluation metrics (hallucination tracking, latency reduction, cost optimization) for generative AI.
Training & Enablement Before deployment, the FDE will complete an intensive 12-hour hands-on coding bootcamp led by Google AI SMEs on the GECX / G3 stack, mastering deterministic fallbacks, rapid debugging, and account turnaround protocols.
Sign in
To continue your applicationor continue with email
By continuing you agree to our Terms & Privacy Policy.
