Honeywell looking for someone who has Intel Loihi or Brainchip Akida experience.
(7) Artificial Intelligence & Machine Learning Systems Engineer | Honeywell | LinkedIn
Extract :- ā
About the job
Job Description
Weāre seeking a highly skilled Artificial Intelligence & Machine Learning Systems Engineer to architect, design, and develop advanced AI/ML systems that power our next generation of products. In this leadership role, youāll contribute to the technical roadmap, mentor engineering teams, and collaborate with cross-functional teams to deliver intelligent, scalable, and production-ready AI and machine learning technologies. You will be responsible for researching, creating, adapting and evaluating AI/ML techniques to solve complex customer problems with real-time solutions to support our defense customers.
Specifically, we are building next-generation cognitive electronic warfare systems that operate autonomously at the tactical edge in contested, low-SWaP (Size, Weight, and Power), denied, and disconnected environments. This is not a prompt-engineering or GenAI role. We are looking for hardcore AI/ML systems engineers who treat machine learning as a component of a larger, mission-critical, real-time embedded system.
Major Duties & Responsibilities
- Design, implement, and harden on-line and continual-learning ML algorithms for RF signal classification, adaptive jamming, cognitive radar, and electronic attack/support decision engines.
- Port, optimize, and deploy ML inference algorithms to edge processors.
- Build and maintain low-latency, deterministic inference pipelines that integrate tightly with real-time RF front-ends and digital signal processing chains.
- Lead the systems integration of AI/ML techniques into mission-critical embedded platforms running real-time operating systems.
- Design and deliver warfighter-focused engineering visualizations and tactical displays (real-time spectrum awareness, threat emitter tracks, cognitive EW decision overlays, confidence heatmaps) using modern web stack frameworks that run natively on embedded tactical processors and dismounted soldier systems.
- Own the MLOps and DevSecOps pipeline for classified EW programs: secure CI/CD, model versioning, containerized build/test/deploy, SBOM generation, and compliance with DoD zero-trust and CNCF security standards.
- Architect and deploy Kubernetes-based edge orchestration clusters (e.g. k3s) that operate in fully air-gapped tactical environments with strict latency and availability requirements.
- Perform end-to-end performance profiling (memory bandwidth, cache coherency, DMA, GPU/TPU/NPU utilization).
- Review code, guide architecture decisions, and mentor the AI/ML engineering team.
- Collaborate with product and engineering teams to identify AI/ML-driven opportunities.
Why This Role Is Different
- You will own the entire stack from algorithm research to bare-metal deployment on platforms that fly, float, or roll into harmās way
- No Python notebooks in production, everything is compiled, containerized, signed, and deployed with cryptographic integrity
- Real impact: your code will out-think and out-maneuver adversary emitters in real conflicts. If you live for the intersection of cutting-edge machine learning and extreme systems engineering under the harshest constraints, we want to talk to you
Qualifications
Required Qualifications:
- Bachelorās in Computer Science, Machine Learning, Artificial Intelligence, Data Science, or related field
- 7 plus years of professional experience shipping production AI/ML systems, ideally in defense, aerospace, or autonomous systems
- Prior work on DoD cognitive EW programs
- Deep expertise in high-performance and real-time applications (not just scripting wrappers)
- Real-time and embedded application programming (no Python-only backgrounds)
- Proven track record of deploying AI/ML solutions to cloud and edge/constrained devices
- Strong systems engineering background: you understand clocks, interrupts, DMA, cache hierarchies, memory-mapped I/O, and real-time scheduling
- Hands-on experience building and securing CI/CD pipelines for classified or regulated environments
- Expertise with Docker, container hardening, and Kubernetes in disconnected/edge configurations (k3s, microk8s, Rancher Harvester).
- Familiarity with RF/ML intersections: signal detection & classification, modulation recognition, emitter geolocation, fingerprinting, adaptive waveform design, or reinforcement learning for EW
- Proficiency with ML algorithms (including NLP, Computer Vision, time-series), libraries including foundational understanding and expertise in statistics probability theory and linear algebra
- Strong understanding of machine learning fundamentals: supervised/unsupervised learning, deep learning, model evaluation, optimization, feature engineering, etc
- Experience with data engineering workflows and building robust training datasets
Preferred Qualifications
- Masterās degree in Computer Science, Machine Learning, Artificial Intelligence, Data Science, or related field
- Experience as the technical lead for establishing and accrediting classified AI/ML information systems under the DoD Risk Management Framework (RMF):
- Author and maintain System Security Plans (SSP), Security CONOPS, and AI/ML-specific risk annexes
- Build and harden multi-enclave classified development, integration, and operational environments (RHEL 8/9, SELinux enforcing, DISA STIGs, Assured Compliance Assessment Solution (ACAS))
- Lead the creation of AI/ML-specific artifacts for eMASS packages, including model cards, data provenance, adversarial robustness testing, and continuous monitoring plans
- Obtain and maintain Authority to Operate (ATO) for classified cognitive EW systems containing advanced GPU/NPU-accelerated AI infrastructure
- Perform Linux systems administration at the classified level: kernel tuning for real-time determinism, custom security hardening, cross-domain solution integration, auditd/ELK stack management, and FIPS 140-3 compliant cryptography
- Deep Linux systems administration and hardening experience in classified environments (RHEL/CentOS, STIG compliance, SELinux policy authoring).
- Hands-on experience authoring RMF packages and obtaining ATOs for systems containing machine learning components for the U.S. Government (Army, Navy, Air Force, or IC customer)
- Expertise with Docker, container hardening (CIS, OSCAP), and Kubernetes in disconnected tactical environments
- Experience or exposure with implementing Government reference architectures
- Experience with neuromorphic or spiking neural network hardware (Intel Loihi, BrainChip Akida)
- Experience with distributed training, GPU acceleration, and high-performance ML compute
- Strong background in foundation algorithms, transformers, or multimodal AI
- Knowledge of automated model monitoring, drift detection, and lifecycle management
- Experience integrating ML models into consumer or enterprise products
That's a crazy amount of experience/qualifications they demand, but nice that Honeywell want someone with experience in neuromorphic or spiking neural network hardware.
Sadly, it also shows the long timeframes we are talking about, they want to architect/design/develop their next generation of products.
From some AI boot: The timeline for electronic product development varies based on factors such as product complexity, regulatory compliance, and design iterations. On average, the process can take anywhere from six months to two years, depending on the number of prototyping cycles, testing phases, and manufacturing readiness.
Explore the complete electronic product development process, from concept to launch, in this comprehensive, practical guide for engineers and innovators.
www.vergentproducts.com
Further, Akida2 only exists in the cloud, as we can see at BRN CES2026. I am not sure whether that will work for all testing, I can imagine they want some real physical chips they can build into their own prototype.
Sean was talking about getting AKD2 made up, Steven B. was also talking about customers want some chips in physical form for testing before committing!
Better get that AKD2 made up in some standard form widely used, the clock is ticking!!
BrainChip at CES 2026 BrainChip demonstrated Akida-powered edge AI in real products at CES 2026. This was a great opportunity to see the technology in action, connect with the team, and explore what the roadmap means for your next device. Live Partner Demos Ultra-Low
brainchip.com
The global microprocessor market size is projected to grow from $130.27 billion in 2026 to $200.76 billion by 2034 at a CAGR of 5.56% during the forecast period
www.fortunebusinessinsights.com
as this company will be the leaders 200 %. On their website they have video demonstrations which are out if this world as far as advanced technologies.
