Job ID:
274429
Published date:
17/09/2026
Summary
Chief AI Prototyping Researcher
Location: Austin, Texas Role-type: Full-time, reporting to the Director of AI
Mission
1AU Technologies is developing microwave and laser technologies for commercial and government applications. This role turns uncertain technical questions into clear prototype opportunities, working demonstrations, and evidence that leaders can use to decide what to pursue next. You will use AI to research unfamiliar problem spaces, define use cases, test assumptions, and move promising ideas from concept to an executable plan.
You will build agentic workflows that search and synthesize technical literature, patents, market information, program needs, simulation results, and experimental data. You will then combine that research with rapid software and hardware prototyping. The work spans machine learning, physics, signal processing, systems engineering, and product discovery. Success requires a generalist who learns quickly, works closely with subject-matter experts, and knows when evidence is strong enough to advance a concept or stop it.
Why this role matters
Microwave and laser technologies can support sensing, communications, inspection, manufacturing, autonomy, aerospace, and national security missions. Early opportunities are often difficult to evaluate because the relevant evidence is fragmented across disciplines and the customer need is still forming. This role shortens that discovery cycle by using agentic research, simulation, and disciplined experimentation to expose technical feasibility, user value, and transition risk before the organization makes a large commitment.
What success looks like in 12 to 18 months
- A repeatable agentic research workflow is operating: Technical and market questions are decomposed, sourced, challenged, and converted into concise opportunity assessments with traceable evidence.
- A prioritized prototype portfolio is in place: Candidate concepts are ranked against mission or customer need, technical feasibility, differentiation, development cost, security constraints, and a clear learning objective.
- Software and hardware prototypes have been delivered: Microwave and laser concepts are demonstrated through working prototypes or high-fidelity simulations that produce quantitative evidence against defined commercial and government use cases.
- Agentic tools are improving research speed and quality: Reusable agents help with literature review, data analysis, code generation, simulation, experiment planning, and technical critique while preserving human review and source provenance.
- Promising prototypes are ready to transition: Each mature concept has a documented architecture, performance envelope, data package, technical risks, validation plan, and recommended path to a customer demonstration, program, or product roadmap.
- A durable knowledge base is available: Research findings, negative results, data, code, assumptions, and decisions can be reused by engineering, product, business development, and program teams.
Core responsibilities
1. Research and scope high value prototype opportunities
- Translate commercial customer problems and government mission needs into specific technical questions, prototype objectives, measurable success criteria, and short validation plans.
- Use agentic workflows to search and synthesize papers, patents, standards, public program information, market signals, supplier capabilities, and internal technical data. Record sources and distinguish verified evidence from hypotheses.
- Assess candidate microwave and laser applications for feasibility, differentiation, transition path, data availability, development burden, security constraints, and potential value.
- Produce concise concept briefs that explain the use case, proposed system, enabling AI, competing approaches, unknowns, experiment plan, and decision criteria.
- Maintain a balanced portfolio of near-term prototypes and higher-risk concepts, and recommend when to advance, redirect, partner, or stop work.
2. Build reliable agentic research and prototyping workflows
- Design agentic workflows that plan research, retrieve evidence, use technical tools, generate and test code, run analyses or simulations, critique outputs, and synthesize findings for human review.
- Connect agents to approved knowledge sources, repositories, data stores, modeling tools, and test infrastructure with clear permissions, provenance, and reproducibility.
- Develop evaluations and human checkpoints that detect unsupported claims, faulty code, weak sources, prompt injection, data leakage, and misleading technical conclusions.
- Operate workflows in environments appropriate to the data and customer, including separation between public, proprietary, export-controlled, and classified work when required.
3. Develop and evaluate microwave and laser prototypes
- Build rapid software and hardware prototypes that combine AI, simulation, signal processing, optimization, and lightweight interfaces to test the core technical claim.
- Work with microwave and radio-frequency experts on concepts involving sensing, radar, communications, electronic warfare, inspection, power delivery, and spectrum operations.
- Work with laser and photonics experts on concepts involving optical sensing, communications, metrology, beam control, manufacturing, autonomy, and government mission systems.
- Design experiments, hardware-in-the-loop tests, and data analyses that quantify performance, uncertainty, failure modes, scaling limits, and operational relevance.
4. Create practical tools and data foundations
- Deliver notebooks, applications, APIs, dashboards, and AI assistants that let researchers and engineers explore results, compare concepts, and reproduce key analyses.
- Build data pipelines for experimental measurements, sensor data, simulation output, imagery, radio-frequency data, documents, and open-source information. Preserve metadata and traceability.
- Create reusable components for agent orchestration, retrieval, evaluation, optimization, experiment tracking, and multimodal data exploitation so each prototype does not start from zero.
- Write maintainable code, tests, and documentation, and collaborate with infrastructure and security partners when a prototype needs to move into a controlled or production environment.
5. Transition prototypes into programs products and intellectual property
- Communicate findings to technical and nontechnical decision-makers, including what is known, what remains uncertain, and which next experiment will reduce the most important risk.
- Partner with engineering, product, business development, and government program teams to shape demonstrations, proposals, development roadmaps, and customer engagements.
- Document inventions, methods, assumptions, negative results, model limitations, and design decisions. Support patent disclosures, technical reports, and publications when appropriate.
Qualifications
- Advanced degree in computer science, electrical engineering, physics, applied mathematics, data science, or a related field, or equivalent applied research experience.
- Five or more years applying AI, machine learning, optimization, or data science to real technical systems, from ambiguous question through tested prototype.
- Broad AI research experience in several areas, such as agentic systems, computer vision, object recognition, signal analysis, reinforcement learning, optimization, or multimodal learning.
- Hands-on experience building agentic workflows with large language models, retrieval, tool use, code execution, planning, evaluation, and human oversight.
- Strong rapid-prototyping judgment: identify the smallest credible test, build it quickly, measure it honestly, and change direction when the evidence does not support the idea.
- Strong Python and software engineering skills, including modern machine-learning frameworks, scientific computing, APIs, version control, and reproducible experiments.
- Ability to synthesize technical, market, and mission evidence, identify gaps or conflicting claims, and convert research into a scoped prototype plan.
- Experience working with large, messy, or multimodal data sets and selecting methods that fit the data, operational constraints, and decision being made.
- Clear communication with researchers, engineers, executives, customers, and government stakeholders.
- Ability to obtain and maintain a U.S. government security clearance, up to TS//SCI, and to comply with applicable security and export-control requirements.
Nice to have
- Experience supporting U.S. national security, defense, intelligence, or government research and development programs.
- Knowledge of microwave or radio-frequency systems, including radar, communications, sensing, spectrum operations, electronic warfare, antennas, or high-power microwave technologies.
- Knowledge of lasers, photonics, optical systems, beam control, electro-optical sensing, semiconductor lasers, or high-power laser technologies.
- Experience with laboratory instrumentation, hardware-in-the-loop testing, embedded systems, signal processing, or rapid hardware integration.
- Experience with electromagnetic, optical, thermal, circuit, or multiphysics simulation connected to experimental data.
- Experience with cloud or high-performance computing, large-scale data processing, MLOps, or restricted computing environments.
- A current U.S. government security clearance, relevant patents or publications, or a record of transitioning prototypes into funded programs or products.
Leadership traits we value
- Curious generalist: You learn unfamiliar technical domains quickly, ask precise questions, and know when to bring in a specialist. You form useful mental models without pretending to replace expert judgment.
- Rapid builder: You choose prototype fidelity based on the decision at hand, instrument the result, and keep the code and evidence organized enough for others to reproduce and extend the work.
- Grounded in evidence: You challenge attractive ideas, verify agent outputs, report uncertainty and negative results, and state what new data would change your conclusion.
- Systems minded collaborator: You connect customer needs, physics, hardware, software, AI methods, security, and program constraints. You explain the tradeoffs in language each discipline can use.
- Security minded: You treat data handling, access controls, provenance, export controls, and operational security as design requirements. You choose tools and environments that fit the sensitivity of the work.
- Focused on users and missions: You spend time with stakeholders, convert their needs into measurable tests, and judge a prototype by whether it answers a real customer or mission question.
- Comfortable with ambiguity: You define the problem while solving it, make reversible decisions with incomplete data, and revise priorities as evidence changes. You keep the team moving without hiding uncertainty.