Lacks experience with large scale systems; background is primarily in embedded systems.
First, the candidate's production experience is primarily in RAG/LLM application engineering, with agentic work limited to project-level efforts. The client is specifically seeking individuals with substantial hands-on experience building 0 to 1 production agentic systems, and has previously passed…
Need someone with extensive hands-on experience building agentic AI systems, including deep judgment on architecture, reasoning, evaluation, failure modes, and solving ambiguous problems independently.
Has a very surface-level understanding of AI and lacks the foundational engineering experience required for this role.
Tweets, press, and people that show why this team is worth your time.
Lexor is hiring its first cohort of Forward Deployed Engineers, split roughly 50/50 between two teams. The role is a Member of Technical Staff (MTS) position spanning applied AI research and production systems, with direct partner-facing responsibilities to ensure operational success inside US Government partner environments.
Two teams under the same role framing:
- Team 1 — Black-Box / Top-Secret Team: works on non-civilian agencies (Department of War, Homeland Security). US citizens only, non-negotiable. Engineers on this team will work on-site at the Pentagon 2-3 times per week and will be sponsored for TS clearance. Existing TS clearance is a strong plus and a real acceleration
- Team 2 — Civilian Agency Team: works with CMS, HHS, state and local governments. US citizens preferred but not required (O-1 candidates are on the table).
The mandate:
- Architect and ship agentic AI systems for high-stakes US Government missions, from proof-of-concept through production
- Design and integrate LLM-based applications — multi-agent frameworks, tool-calling agents, RAG pipelines, human-in-the-loop autonomy workflows
- Embed directly with government partners in operational environments
- Build and scale data curation and evaluation pipelines for model training and iteration
- Develop and maintain infrastructure supporting model inference, experimentation, evaluation, and deployment on frontier AI platforms
- Turn mission-relevant data, models, and tooling into real operational autonomy capabilities
- Own the full system lifecycle — discovery, architecture, deployment, security, reliability, continuous improvement
Engineers are expected to be proficient at expressing technical ideas clearly to non-technical government stakeholders, and to represent Lexor at a high caliber
Requirements
- 1-4 years of production engineering experience
- Strong Python engineering background with demonstrated end-to-end delivery
- Hands-on experience with LLMs and building agentic systems — RAG pipelines, tool-calling agents, multi-agent frameworks, context management, agentic engineering
- Real deep technical knowledge, not vibe-coding
- Track record designing or managing data pipelines and ML infrastructure
- Excellent verbal communicator
- In-person 5 days/week at DC office
- US citizen for the top-secret team (non-negotiable); US citizen preferred for the civilian team, O-1 candidates on the table for exceptional profiles
- Available to start ASAP
Profile Tier 1 — Ex-Founders / Ex-Founding Engineers of AI Startups Ex-founding engineer experience, having launched a startup — really would hit all the criteria Especially: ex-YC founders, Neo Scholars, Z Fellows, South Park Commons, Pioneer, a16z Speedrun alumni
Tier 2 — Forward Deployed / Applied Engineers at Defense-Adjacent Palantir Foundry FDE / delivery engineering Anduril, Shield AI, Rebellion Defense, Vannevar Labs, HavocAI Scale AI Federal / Applied Skyryse, Applied Intuition
Tier 3 — Applied Engineers at Agentic AI / LLM Companies Sierra, Decagon, Glean, Cresta, Rox, Parloa OpenAI Solutions, Anthropic Applied Perplexity Applied, Cursor Applied, Cognition (Devin) Applied Together AI, Fireworks AI, Modal, Replicate — applied side
Tier 5 — Ex-DOGE / Ex-USDS / Ex-18F Engineers Engineers who did tours in DOGE (2025), US Digital Service, 18F, or the White House Presidential Innovation Fellows Direct experience with government data systems, procurement rhythms, security constraints
Tier 6 — Big Tech with Zero-to-One Signal (Narrow) Big-tech candidates need visible startup founding, hackathon wins, OSS contributions
Strong Bonuses
- Active US security clearance (Secret, Top Secret, TS/SCI, TS/SCI with polygraph)
- Prior work on US Government, defense, intelligence, or national security programs
- Prior work on Medicare / Medicaid data, CMS systems, healthcare fraud detection, or public-sector fraud analytics
- Fine-tuning + eval track record on proprietary datasets
- Reinforcement learning workflows or simulation-based environments (JD explicit)
Location and Visa
- DC office 5 days/week in-person
- US citizen only for top-secret team
- US citizen preferred for civilian team