What are the responsibilities and job description for the Founding AI Engineer position at Atem Corp?
Title: Founding AI Engineer / Member of Technical Staff (YC-backed public-safety startup)
Location: NYC (Preferred) or SF Hybrid
Must-Have Requirements
Location: NYC (Preferred) or SF Hybrid
Must-Have Requirements
- 3 years of professional software engineering experience, with strong backend fundamentals (distributed systems, APIs, data modeling) in a modern stack (e.g., Python TypeScript/React or similar).
- Hands-on experience building and shipping ML/AI systems, ideally with LLMs or other deep-learning models used by real users (not just research or prototypes).
- Experience with retrieval / RAG or similar pipelines over unstructured text or multi-modal data (documents, audio transcripts, etc.), including designing data flows and evaluation approaches. (Based directly on the JD's "Data Versatility / RAG or similar pipelines" section.)
- Comfort working in a fast-moving startup environment with high autonomy, ambiguity, and end-to-end ownership of projects.
- Strong communication skills able to partner with founders, FDEs and investigators, explain technical trade-offs, and turn messy requirements into robust systems.
- Responsibilities: Owning Core AI and Backend Systems: Design and manage ML systems for ingesting and analyzing large evidence volumes. Data Retrieval Architecture: Create reliable retrieval/RAG pipelines for quick investigator access to structured and unstructured data. Model Prototyping: Experiment with LLMs and other novel tools, developing robust production systems for law enforcement. User Collaboration: Implement ML features using feedback from detectives and prosecutors, enhancing...
- This is a founding AI / backend engineering role. You'll design and ship the ML systems that power Closure's "digital analyst" for law enforcement-working closely with the founders, Forward-Deployed Engineers, and investigators in the field.
- What you will do:
- Own core AI and backend systems that ingest, process, and search across large volumes of evidence (calls, reports, documents, transcripts, and more).
- Design and implement retrieval / RAG pipelines for unstructured and structured data, making it fast and reliable for investigators to find what they need.
- Prototype with new models and tools (LLMs, embeddings, vector databases, observability stack), then harden the best ideas into production systems agencies can trust.
- Collaborate closely with Forward-Deployed Engineers and users to turn real-world feedback from detectives and prosecutors into concrete ML features and ranking improvements.
- Contribute across the stack when needed (APIs, internal tools, evaluation dashboards) to keep the overall AI surface area robust, monitored, and maintainable.
- 3 years of professional software engineering experience with strong backend fundamentals (distributed systems, APIs, data modeling) in a modern stack (e.g., Python TypeScript/React or similar).
- Hands-on experience building and shipping ML/AI systems used by real users, ideally involving LLMs or other deep-learning models (not just research or PoCs).
- Experience with retrieval / RAG or similar architectures over unstructured text or multi-modal data (documents, transcripts, logs), including designing data pipelines and evaluation approaches.
- Comfortable working end-to-end: from understanding investigator workflows and problem framing, to designing experiments, to deploying and monitoring models in production.
- Strong communication and collaboration skills; able to work directly with founders, Forward-Deployed Engineers, and non-technical stakeholders in a small, fast-moving, mission-driven team.
- Field-Driven Engineer Strong full-stack engineer (Python modern frontend) who enjoys leaving the office, sitting with users, and seeing how software actually gets used in the wild.
- Customer-Obsessed Problem Solver Comfortable building trust with detectives and agency leadership, asking good questions, and turning messy requirements into clear product and technical decisions.
- High-Ownership Operator Thrives in tiny, fast-moving teams, takes full responsibility for deployments and outcomes, and is happy to do whatever the situation requires (from debugging to running training sessions).
- Mission-Motivated Energized by improving public safety and the criminal-justice system, and comfortable working with sensitive, sometimes difficult case material.
- Startup-Ready Has prior experience in early-stage or talent-dense environments and is excited by ambiguity, rapid iteration, and having a big say in how the product and company evolve.