AI Engineer

Aslan; Washington (District of Columbia); On-site; USD 150,000 - 210,000; Full time

Job summary

Aslan is hiring an AI engineer to build the cognitive architecture and orchestration for autonomous agents operating over extended horizons in high‑stakes environments. You’ll push frontier models with post‑training, domain‑specific fine‑tuning, and scaffolding to enable robust multi‑agent collaboration and decision making.

You’ll own data feedback and evaluation pipelines, ensuring reliability, coherence, and operational impact while working with platform engineers to deliver mission‑critical

Qualifications

• Roughly 5+ years of software/ML engineering experience, at a senior level of judgment and autonomy.
• Hands‑on experience building agentic AI systems that operate autonomously over extended time.
• Deep familiarity with frontier models: fine‑tuning, post‑training, and domain adaptation.
• Strong software engineering and systems thinking with production‑grade tools.
• Rigorous data‑oriented mindset: provenance, labeling, and evaluation as core engineering concerns.
• Experience running models on constrained hardware: quantization, memory management, inference optimization.

Responsibilities

• Design the cognitive architecture enabling AI agents to operate autonomously over extended task horizons.
• Build orchestration for multi‑agent coordination at scale with context sharing and conflict resolution.
• Advance frontier models via post‑training, domain‑specific fine‑tuning, and scaffolding strategies.
• Develop evaluation frameworks for long‑duration deployments: reliability, coherence, regression detection.
• Own data feedback loops from live deployments to drive model improvements while respecting domain constraints.
• Collaborate with forward‑deployed and platform engineers to operationalize requirements.

Skills

Senior software/ML engineering

Agentic AI systems

Frontier model understanding

Python and PyTorch

Data provenance & evaluation

Constrained hardware optimization

Tools

Python

PyTorch

Job description

About Aslan

Our national security apparatus was designed for an adversary that congregated and operate in the physical world. The threats disrupting our way of life today have gone faceless, borderless, and beyond the reach of any human operator. Aslan builds the autonomous agents that reach them, unmask them, and stop them, before a life can be lost, a household can be bankrupted, innocence can be stolen, or our can be nation subverted.

In 12 months we've gone from founding to live operational tasking and pilots across multiple national security and law enforcement partners. We've proven it works. This role makes it permanent.

We’ve raised $20M to date from Khosla Ventures, XYZ, BoxGroup, 2048, Liquid 2, Precursor Ventures, and others.

Role

We're not just repackaging commodity capabilities and building workflows. Aslan is pushing the boundaries what agents can accomplish for the highest stakes national security priorities. To remove the bottlenecks that keep us outnumbered and outmaneuvered by digital adversaries, we're hiring an AI engineer that lives and breathes the frontier of agent scaffolding, orchestration, and multi-agent systems.

You'll build the cognitive architecture that enables AI agents to force multiply their human counterparts, taking domain-specific autonomy beyond what off-the-shelf models can robustly accomplish. You'll figure out how frontier models can be used in heterodox ways and deploy the cutting-edge in post-training to take them in directions that were never previously imagined. Most importantly, you'll go beyond mere clever chaining and compound systems to build foundational improvements in task horizon to give the operators on the frontlines the leverage they deserve and the scale they need.

Responsibilities

• Design and build the cognitive architecture that lets Aslan's agents operate autonomously over extended task horizons: planning, reasoning, memory, and decision‑making systems that sustain coherent, reliable behavior across days and weeks in adversarial environments.

• Build the orchestration layer for multi‑agent coordination at scale, including context‑sharing, conflict resolution, and behavioral consistency across concurrent operations.

• Push frontier models beyond their default capabilities through novel post‑training, domain‑specific fine‑tuning (SFT, preference optimization, LoRA/QLoRA), and scaffolding strategies purpose‑built for high‑stakes autonomous operation.

• Build evaluation frameworks for agentic behavior over extended deployments, focused on operational reliability, behavioral coherence, and regression detection at the timescales our agents actually operate.

• Own the data feedback loop from live deployments, turning raw operational signal into training data and evaluation benchmarks that compound model improvements over time, under the sensitivity constraints of the domain.

• Partner closely with forward‑deployed and platform engineers to understand what “better” means operationally and get your systems into the hands of the people running missions.

What You’ll Bring

• Roughly 5+ years of software/ML engineering experience, at a senior level of judgment and autonomy. We’re flexible on the senior‑vs‑staff label; we care about depth.

• Hands‑on experience building agentic AI systems that operate autonomously over extended time. You’ve designed the scaffolding, orchestration, or memory systems that let agents do sustained, real work.

• Deep familiarity with how frontier models behave, fail, and get extended. You’ve fine‑tuned open‑weight LLMs, pushed them into novel territory through post‑training or architectural changes, and shipped the result.

• Strong software engineering and systems thinking. You build reliable, observable, debuggable production systems. Python, PyTorch, and the modern ML stack.

• Rigorous instincts on the data side. You know model quality is mostly a data problem and you treat provenance, labeling, and evaluation as first‑class engineering concerns.

• Working knowledge of running models on constrained hardware: quantization, GPU memory management, inference optimization.

Bonus Points

• Experience building agents that operate in adversarial or contested environments where robustness to unexpected inputs and graceful degradation actually matter.

• Prior work with multi‑modal systems (text, image, audio) in an agentic context.

• Experience with air‑gapped or classified environments and their deployment constraints.

• Active, recent, or eligibility for a U.S security clearance (TS preferred).