Chief Software Architect

Description Role Overview

We are seeking a Chief Software Architect to serve as the technical lead and systems-level architect for our spacecraft autonomy stack.

This role will define and own the end-to-end software architecture for autonomy capabilities supporting SDA and ISAM missions:

• Multi-sensor perception and world modeling
• Estimation and relative navigation support for ISAM/RPO missions
• Data pipelines (synthetic + real)
• Embedded deployment to ground station infrastructure, as well as flight/edge hardware
• Government-ready security posture and delivery workflows
• Operational reliability and safety engineering

You will lead a team of engineers spanning Computer Vision, Full Stack, Dev Sec Ops /Platform, Mod/Sim, and Embedded/Edge deployment
, while working closely with company leadership, mission/operations teams, and hardware engineers.

This is a role for someone who wants to architect autonomy that flies — not just publish papers or create prototypes that never leave the lab.

• Define and maintain the end-to-end architecture for a production-grade spacecraft autonomy stack, spanning:
  • perception + sensor ingest (EO/IR, multi-camera, star trackers, etc.)
  • calibration, time-sync, and sensor health monitoring
  • world modeling / representation (relative state, scene understanding, object modeling)
  • state estimation interfaces (relative navigation, uncertainty propagation)
  • decision & autonomy interfaces (planning hooks, constraints, safety gating)
  • fault detection, fallback modes, and confidence-driven behavior switching
• Build scalable, modular designs that transition cleanly from:
  • R&D prototypes → validated algorithms → flight/edge-capable products → mission operations
  • including clear separation between:
    • offline training/evaluation pipelines
    • on-orbit/edge inference pipelines
    • mission operations tooling
• Establish rigorous interface contracts and architectural guardrails so autonomy capabilities remain:
  • reliable
  • testable
  • upgradeable
  • safe to deploy into mission environments (especially for ISAM / RPOD scenarios)

• Establish patterns for high-reliability autonomy software:
  • modular architectures, clear interfaces, versioning
  • deterministic execution where needed
  • fault detection, fallback modes, and observability
• Implement rigorous engineering practices without slowing innovation.

• Drive design decisions across the entire compute surface area:
  • cloud data pipelines and model training
  • simulation environments and evaluation infrastructure
  • embedded inference on real hardware (e.g., GPU edge devices / flight-relevant compute)
• Ensure production performance in real constraints:
  • latency, memory, power, bandwidth, thermal constraints
  • repeatability, robustness, recoverability

• Work with Dev Sec Ops  to ensure architecture supports:
  • secure builds, controlled release pipelines
  • Government security requirements, traceability, auditability
  • secure deployment into customer/government environments CUI up to TS levels
• Design for safety in autonomy contexts:
  • bounded behaviors
  • explicit failure modes
  • confidence estimation and gating

• Lead and mentor a multidisciplinary autonomy software team (~10+ engineers in year 1, and multiple teams in year
  2)
• Conduct design reviews, set coding standards, define architecture guardrails
• Help recruit top-tier engineers and develop internal technical leadership

• Stay current on the autonomy and perception research landscape:
  • CV foundation models, pose estimation, multi-view geometry
  • neural implicit representations, 3D reconstruction
  • tracking/filtering methods, uncertainty quantification
• Translate research into deployable product capability:
  • evaluate, prototype, validate, productize

• Own perception stack architecture across software layers and teams
• Drive end-to-end technical strategy, with clear tradeoffs and rationale
• Create architecture artifacts:
  • system design documents, interface contracts, data schemas
  • “golden path” workflows for training → evaluation…