Forward Deployed Engineer

The Mission

Wall Street runs on code that predates the smartphone. Beneath the gleaming towers of Manhattan lie millions of lines of COBOL, decades-old Java monoliths, and mainframe systems that process trillions of dollars daily. These systems work—but they're strangling innovation. You will be Retrain's tip of the spear in the financial sector. Your mission is to prove that AI agents can do what armies of contractors couldn't: systematically modernize these systems while they're still running. You won't just write code—you'll reshape how the world's largest financial institutions think about software development. This is not a cushy consulting gig. You will be dropped into high-stakes environments where a bad deploy can make headlines. You'll work alongside battle-hardened engineers who've seen every vendor promise fail. Your job is to earn their trust by shipping results they thought were impossible. The banks that master agentic engineering will define the next era of finance. You'll be the one who shows them how.

Responsibilities

• Embed directly with client engineering teams for 3-6 month engagements, working on-site at major banks, hedge funds, and insurance companies across Manhattan.
• Conduct deep-dive audits of legacy codebases spanning COBOL, C++, Java, and proprietary languages—some with minimal documentation and original authors long gone.
• Design and execute 'Strangler Fig' migration strategies using our Agent Swarm to incrementally replace legacy components without disrupting production systems.
• Customize and fine-tune Retrain's core agent models for financial domain-specific tasks: trade reconciliation, risk calculations, regulatory reporting, and compliance checks.
• Build custom MCP (Model Context Protocol) servers that connect our agents to proprietary trading systems, risk engines, and internal databases.
• Train client engineers—from junior developers to principal architects—on agentic development patterns, prompt engineering, and human-in-the-loop workflows.
• Navigate complex stakeholder environments: translate technical progress into business outcomes for C-suite executives, risk committees, and board presentations.
• Maintain rigorous security practices: work within air-gapped environments, handle sensitive financial data appropriately, and ensure all agent actions are auditable.
• Document edge cases, failure modes, and domain-specific challenges; feed these insights back to our Core R&D team to improve the platform.
• Build lasting relationships with client teams that turn into multi-year partnerships and expansion opportunities.

Requirements

• 7+ years of software engineering experience with significant time spent in large-scale, complex codebases.
• Polyglot engineering background: you've shipped production code in at least 4 languages and can pick up new ones in days, not weeks.
• Deep understanding of distributed systems: you can whiteboard CAP theorem tradeoffs, debug race conditions, and design for failure.
• Experience with legacy modernization: you've migrated systems, worked with mainframes, or led large-scale refactoring efforts.
• Fearlessness in the face of messy code: 10-year-old unmaintained systems with zero tests don't scare you—they excite you.
• Exceptional communication skills: you can explain complex AI concepts to skeptical non-technical stakeholders and make them believers.
• Experience in financial services, fintech, or other regulated industries is strongly preferred.
• Comfort with ambiguity: you thrive when the path forward isn't clear and you need to figure it out yourself.
• Willingness to travel up to 25% for client engagements and team gatherings.
• Based in or willing to relocate to the New York metropolitan area.