
Job Description
Senior Backend Engineer - Agents
Location: Remote (US) or Hybrid (Boston)
About Us
Nebulock is an agentic threat hunting platform that autonomously surfaces behaviors, not just IOCs, from various data sources. Nebulock acts like a teammate: a 24/7 AI threat hunter that investigates hypotheses, reasons through telemetry, and learns from an environment. Today, threat hunting is broken. Security teams spend weeks chasing alerts, writing detections by hand, and manually validating findings often just to confirm what their existing tools already flagged. Meanwhile, attackers exploit credentials, move laterally, and operate in silence. Nebulock flips the model. We continuously and autonomously hunt across endpoint, identity, and cloud telemetry. We identify the subtle behavioral signals that point to credential misuse, lateral movement, insider threats, and post-access activity. Then we turn those hunts into hardened, behavior-based detections automatically.
Position Overview
As a Senior Backend Engineer - Agents at Nebulock, you will work on the agent infrastructure and platforms upon which our agents are built, as well as the agents that power our product. You’ll co-own our TRACE Graph and other knowledge stores, ensuring our agents continuously adapt to feedback and learn. This role is ideal for engineers who are relentless about iterating and improving upon non-deterministic, agentic systems. This role will collaborate closely with our detection engineers and threat hunters to ensure that we are encoding their domain expertise into our agents.
Expected Impact
Continuous iteration, evaluation, and cross-team collaboration to ensure that our agents are behaving like the expert detection engineers and threat hunters on staff at Nebulock
Design, build, and maintain scalable data structures that are optimized for our agents (our TRACE graph, embeddings, semantic search etc.)
Ensure that our agents are behaving safely and as expected
Build and maintain automated evaluations and benchmarks for our agents to ensure that we can iterate quickly while not having regressions
Provide feedback and guidance to other engineers who are working on agents at Nebulock
Step in to work on additional backend systems as needed (we are a startup and flexibility is key to our success)
Qualifications
4+ years software engineering experience with 1+ years building customer-facing agentic systems in production
Strong backend software engineering fundamentals
Proficiency in Python, Java, Go, Rust, or similar
Strong understanding of agentic architectures and their tradeoffs
Experience designing, building, and evaluating production RAG systems
Strong experience with agent evals and agent observability
Experience with either AWS or GCP
Ability to adapt, iterate, and ship quickly
Hunger for growth and the desire to work in a low-ego environment
Able to stay up to date with the newest advancements in the LLM and agent space while maintaining a relentless focus on delivering customer value
Nice to Haves
Cybersecurity experience
Startup experience
Experience with graph databases
Experience evaluating the performances of different LLMs
Compensation
We offer a competitive compensation package comprised of a base salary, equity and comprehensive benefits. The base salary for this role is $180,000 - $220,000. Compensation packages are determined based on a number of factors including but not limited to, role related experience, skillset, and geographic location.
Benefits
Health, dental, and vision coverage
Unlimited PTO
Remote working flexibility
Life Insurance
401k
Optimize Your Resume for This Job
Get a match score and see exactly which keywords you're missing
Job Details
- Category
- Software
- Employment Type
- Full Time
- Location
- United States (Remote)
- Posted
- Compensation
- $180,000 - $220,000 per year
About Nebulock
Nebulock is the first agentic threat hunting platform; autonomously surfacing behaviors, not just IOCs, from your existing data. Nebulock acts like a new teammate: a 24/7 AI threat hunter that investigates hypotheses, reasons through your telemetry, and learns from your environment. Whether you’re a two-person SOC or a global enterprise, we scale your threat hunting—and give your team superpowers. Today, threat hunting is broken. Security teams spend weeks chasing alerts, writing detections by hand, and manually validating findings—often just to confirm what their existing tools already flagged. Meanwhile, attackers exploit credentials, move laterally, and operate in silence. Nebulock flips the model. We continuously and autonomously hunt across endpoint, identity, and cloud telemetry—identifying the subtle behavioral signals that point to credential misuse, lateral movement, insider threats, and post-access activity. Then we turn those hunts into hardened, behavior-based detections—automatically. No new agents No alert regurgitation No workflow disruption Just high-fidelity, explainable findings—delivered directly to your SIEM, API, or Slack.
More Roles at Nebulock





Similar Software Roles



Found this role interesting?