# How Forward Deployed Engineering is done at Ramp — Leo Mehr

## Executive summary

The talk outlines two core principles for Forward Deployed Engineering (FDE): always scoping and scaling with tokens. FDE teams must resist the urge to simply say 'yes' to every request; instead, they must rigorously scope requirements by asking context-gathering questions to ensure the right product is built. To scale this function, the entire FDE pipeline—from initial intake to spec writing—must be automated using AI agents (scaling with tokens). This requires building robust agent harnesses and implementing validation loops using evals, rubrics, and human feedback.

## Key takeaways

- Always Scope First: FDEs must pause before accepting urgent requests (e.g., SAP S/4HANA integrations) to gather context and determine what is truly necessary for the customer's success, rather than simply fulfilling the immediate ask.
- Scale with Tokens: The future of FDE requires automating every stage of the request lifecycle—from gathering context to writing a spec—by replacing manual work with AI agents and model capabilities.
- Hybrid Approach is Key: Effective modern FDE needs both human judgment (for taste and final validation) and agent volume/automation to handle the sheer scale of incoming requests.

## Technical details

- FDE Pipeline Automation: The process involves automating stages like request intake, which was demonstrated using Notion agents. This system goes beyond simple processing by conducting multiple rounds of back-and-forth questioning with the submitter until a detailed specification is ready.
- Agent Development Infrastructure: Building scalable AI pipelines requires an 'agent harness' and rigorous validation processes, including evals, rubrics, and human feedback loops, to ensure output quality at each stage.
- Scope Validation Failure: A critical lesson learned was the importance of validating basic assumptions (e.g., mobile platform requirements) early in the process to prevent wasted effort after significant development time has been invested.

## Practical implications

- Implement structured intake workflows (like Notion) to standardize and automate the initial context-gathering phase of customer requests.
- Treat FDE processes as a pipeline that can be broken down into discrete, automatable stages for agent implementation.
- Prioritize building robust validation layers (evals/rubrics) around AI agents to maintain output quality and prevent 'token maxing slop cannon' outputs.

## Topics

Forward Deployed Engineering, Generative AI, Process Automation, Product Management, Agentic Workflows, Leo Mehr, LinkedIn Profile

Source: https://www.youtube.com/watch?v=ITMXwI6QL6A