What does the work look like day to day?
A forward deployed engineer, or FDE, owns a customer deployment from first conversation to stable production. OpenAI's San Francisco posting lists discovery, technical scoping, system design, the build, and production rollout, all done alongside the customer's own engineers. You write production code across frontend and backend, decide what to cut when time runs short, and turn what worked into tools and playbooks the next team can reuse. Success is measured by whether people actually use the system and whether it changes their workflow.
Most of the hard work is integration. OpenAI's financial services posting has FDEs owning data flows, reliability, monitoring and on-call readiness, and building in sensitive data environments where access controls and data lineage shape the design. They also run evaluation loops against benchmarks specific to the workflow, such as accuracy, latency and false positives. At Salesforce, a director described a typical day as writing agent instructions, testing prompts, working through a difficult integration, and helping the product team shape features.
How is it different from solutions engineering and sales engineering?
A Microsoft writer who has tracked the role describes FDE as a job rooted in software engineering, sitting in a different space from sales engineering, solutions architecture and consulting. The practical test is who writes and ships the production code. Sales engineers mostly support a deal with demos and technical answers. Solutions architects mostly design systems and advise. An FDE stays after the contract is signed and builds the system inside the customer's environment, with real data and real constraints.
The lines blur in practice. OpenAI's financial services FDE leads scoping from pre-sales through post-production, so the job touches the sales cycle. Databricks lists one opening as Senior Forward Deployed Engineer and Solutions Architect in the same title. Salesforce compares its FDEs to traditional tech consultants, but says they move faster, and that partners still handle much of the routine implementation while FDEs bring inside knowledge of the product. The other difference is the feedback loop: FDEs are expected to change the vendor's product roadmap.
Which companies use the title?
Palantir made the role popular, embedding engineers with customers who were mostly government agencies. Salesforce says Palantir called them Deltas and had more of them than regular software engineers until 2016. AI model companies adopted the title next. OpenAI lists FDE roles in San Francisco, New York, Paris, Munich and Washington, DC, and Anthropic's careers page lists FDE and FDE manager roles in New York, London, Munich, Paris, San Francisco and Seattle.
Platform and cloud companies have followed. Salesforce launched its FDE team in April 2025 and has committed to building a team of 1,000. Databricks hires senior FDEs by industry, including retail, manufacturing and public sector. AWS announced a Forward Deployed Engineering organization backed by a $1 billion investment to embed thousands of engineers with customers, and it extends the model to consulting partners. Salesforce cites Indeed and Financial Times analysis showing FDE postings rose more than 800% between January and September 2025.
Is this a lasting career or a passing title?
The title is new in its AI form, but the work underneath it is not. Andrew Ng, writing in DeepLearning.AI's newsletter, links the role's comeback to how much effort it takes to turn a general model into a custom agentic workflow for one business. He expects AI engineering to split into more specialized roles over time, the same way software engineering split into frontend, backend and devops.
OpenAI frames the job as one that keeps making its own past work obsolete: solve a hard problem for one customer, then package it so the next customer needs far less custom work. For a career, that means the lasting skills are production engineering, data integration, evaluation and talking clearly with business stakeholders. Those carry over to AI engineer, solutions architect and engineering manager roles if the FDE title fades.
Pay
USD a year, before tax. US figures. The BLS rows come from BLS OEWS May 2025 for software developers, moved to 2026 with the Employment Cost Index. That is the nearest series: it covers the whole occupation, and BLS has no FDE or AI series. The posting rows are base pay from single job ads at large AI and data companies. Both ads say total pay can also include equity and bonus. Refer to the AI Integration Engineers salary guide for figures from live adverts.
What you need
- Production codingOpenAI asks for production-grade code across frontend and backend in Python, JavaScript or similar stacks. Databricks lists Python, Scala or JavaScript/TypeScript.
- Experience levelOpenAI's FDE postings ask for 5+ years of engineering or technical deployment work that includes customer-facing work. Databricks' senior retail FDE role asks for 6+ years in data engineering, data platforms or software engineering.
- Data and cloud platformsDatabricks wants working knowledge of two or more of AWS, Azure and GCP, deep Apache Spark experience, CI/CD for production, and working knowledge of MLOps and AI APIs.
- Evaluation and LLM deploymentOpenAI's financial services role asks for experience deploying systems powered by LLMs or generative models, and running evaluations against workflow-specific benchmarks.
- Customer communicationFDEs explain trade-offs to executives and non-technical staff. Salesforce ranks problem-solving first, followed by technical range, communication, business sense and a habit of constant learning.
- Willingness to travelOne OpenAI FDE posting says travel of up to 50% may be required. A Databricks senior FDE posting lists customer travel of 20%.
How to get there
- Years 1-3Build software engineering depthShip production code across the stack and learn to own systems in production, including monitoring and on-call. This is the base every FDE posting assumes.
- Years 2-4Add data and AI integration skillsWork on data pipelines, cloud platforms, LLM APIs and evaluation. Build at least one AI system that real users depend on.
- Years 2-5Take on customer-facing workVolunteer for customer escalations, implementations or internal platform rollouts where you must scope messy requirements and explain trade-offs.
- Year 5+Move into an FDE rolePostings at OpenAI and Databricks ask for 5 to 6+ years of experience. Moving internally also works: Salesforce describes a technical architect who joined its FDE team through a six-week onboarding program.
- Year 7+Specialize or leadGrow into staff FDE, FDE manager, or an industry focus such as financial services or public sector, all titles seen in current postings.
Open now on AutomationRoles.ai
Questions people ask
Is a forward deployed engineer the same as a solutions engineer?
No. Solutions and sales engineers mainly support the sale and design the approach. An FDE writes and ships the production system inside the customer's environment, often after the deal closes, and sends field lessons back to the product team.
How much do forward deployed engineers travel?
It depends on the employer. One OpenAI FDE posting says travel of up to 50% may be required, and a Databricks senior FDE posting lists 20%. Salesforce says its FDE pods sometimes travel to the customer and embed in its daily work.
How much experience do FDE roles ask for?
The AI company postings sampled here ask for a lot. OpenAI asks for 5+ years of engineering or technical deployment experience with customer-facing work, and a Databricks senior FDE role asks for 6+ years. Plan on building solid production engineering experience first.
How long does an FDE engagement last?
Salesforce says its FDE pods work full-time with one client for about three months, or until an agent is deployed for one or two use cases. OpenAI gives one example where the first deployment took two months and extending it to a second team took one day.
Is there a certification for forward deployed engineers?
No single certification covers the role, and there is no license or regulator for it. Databricks lists a Databricks certification as a plus for its FDE roles, and AWS requires partner engineers to pass an AWS-defined technical bar before working on its partner-led FDE engagements.
What do forward deployed engineers earn?
The nearest BLS series, software developers, has a 2026-adjusted US median of $140,200, but it covers all developers. Single postings run higher: OpenAI lists a $185K to $300K base plus equity, and a Databricks senior FDE role lists $182,000 to $250,208.
Sources
- OpenAI: Forward Deployed Engineer (FDE), San Francisco job posting
- OpenAI: FDE, Financial Services, New York job posting
- OpenAI: What forward deployed engineering means at OpenAI
- Salesforce: Today's Hottest Role: Forward Deployed Engineer
- Databricks: Sr. Forward Deployed Engineer (FDE), Retail
- Microsoft Command Line: The evolution of forward-deployed engineering
- DeepLearning.AI The Batch: Forward Deployed Engineers and the Future of AI Engineering
- AWS Partner Network Blog: Forward Deployed Engineering for Partners
- Anthropic: Careers job listings
