Forward Deployed Engineer

Position: Forward Deployed Engineer

Skills: LLM APIs (Anthropic, OpenAI, Google), AI coding tools (Cursor, Claude Code, Copilot, or equivalent), Git, basic CI/CD & Docker.

Experience: 0-2 years

Location: Princeton, NJ

Educational Preference: BS or MS in Computer Science, EE, Math, Statistics, Physics, or a comparably rigorous STEM field.

(Candidates with a Computer Science degree/program at US Tier-1/Ivy League universities are highly preferrable)

We at Coforge are hiring Forward Deployed Engineers with the following skillset:

About the team:

• The Coforge FDE Practice is the tip of the spear for our enterprise AI work. Our Forward Deployed Engineers embed with Fortune 500 clients across financial services, insurance, healthcare, travel, and retail to take AI from "interesting pilot" to "production system that moves a business metric." We're scaling the practice over the next year, and we're hiring a founding cohort of Associate FDEs to build that scale-up alongside us.

About the role:

• You'll join a 60~90-day intensive that turns you into a deployable FDE: a foundation in applied AI engineering (LLMs, RAG, agentic systems, evals), a focused vertical immersion, and a capstone built on a client-mirrored problem. You'll be paired with senior FDEs from day one — not lectured at, but pulled into real work the moment you can contribute.

What you'll do after the 90-day academy:

• Embed with an enterprise client team — typically Fortune 500 — to deliver production AI systems in two-to-four-week sprints.
• Co-own a scoped workstream with a senior FDE: agentic workflows, RAG pipelines, eval frameworks, fine-tunes, or AI-native rebuilds of legacy processes.
• Ship production code from day one. Not slideware, not "POCs that never leave Jupyter" — real systems with real users.
• Present your work weekly to client stakeholders, ranging from engineering leads to the C-suite. You'll write the readout. You'll demo the system. You'll defend the design choices.
• Translate ambiguous client problems into shipped solutions — most FDE work begins with a vague pain point, not a written spec.
• Move between clients and verticals as you grow — typical glide path is solo workstream ownership by month nine, leading a deployment by month eighteen.
• Mentor the next cohort of associates as they arrive.

Required Qualifications:

• Education: BS or MS in Computer Science, EE, Math, Statistics, Physics, or a comparably rigorous STEM field. (Candidates with a Computer Science degree/program at US Tier-1/Ivy League universities are highly preferrable)
• Experience: 0–2 years post-graduation, including internships.
• Languages: Production-quality Python. Comfortable reading any modern language.
• Applied AI: Have built non-trivial applications using LLM APIs (Anthropic, OpenAI, Google). Understand RAG, embeddings, vector retrieval, and how evals work.
• AI-native development: Daily user of AI coding tools (Cursor, Claude Code, Copilot, or equivalent) — not occasional, daily.
• Engineering hygiene: Git, basic CI/CD, Docker. You've worked in a shared or large-scale codebase beyond individual projects.
• Written communication: You can write a one-page memo a busy executive will actually read.
• Verbal communication: You can present a technical decision to a non-technical audience. Tested in our interview via a live project walkthrough.

Nice to have:

• Hands-on experience with agentic frameworks (LangGraph, CrewAI, AutoGen, Mastra, OpenAI Agents SDK, MCP).Fine-tuning experience (LoRA, full fine-tune, RLHF basics).Eval and observability tooling (LangSmith, Braintrust, Arize, Weights & Biases)
• Cloud certification or strong hands-on experience with AWS, Azure, or GCP.
• TypeScript or Go alongside Python.
• Public technical footprint: blog, popular GitHub repos, Hugging Face contributions, or active technical writing.
• Prior client-facing experience in any context — consulting, sales engineering, founder, tutoring, debate, teaching.
• Familiarity with one of our focus verticals (banking, insurance, healthcare, travel, retail) — but we explicitly do not require it; we train it.