AI Engineer

Octagon Talent Solutions is a full-service technology recruitment and staffing company based in South Florida. We humanize technical recruitment by identifying candidates' long-term career goals and assessing cultural fit when presenting opportunities. Our focus on long-term relationships helps ensure placements that last.

The opportunity

You’ll help build what many consider a category-defining product: an “industrial operating system” that turns physical-world sensor data into operational decisions. This role is for builders who thrive in a high-ownership, minimal-process environment and can ship quickly with a “mini founder” mindset.

What you’ll build

• Agent runtime + orchestration: tool use, state management, memory, approval gates (human-in-the-loop), tracing/observability, evaluation, and cost controls
• Multimodal pipelines for physical-world data:
• video ingestion + processing (e.g., RTSP)
• computer vision (detection/segmentation/action recognition)
• time-series telemetry (vibration/thermal), FFT feature extraction, cross-sensor correlation
• Context systems grounded in real operations:
• industrial ontology / knowledge graph mapping plants, assets, sensors, work orders, materials, and maintenance history
• memory layer: traces, playbooks, asset templates, and transferable failure-pattern libraries
• Operational decision surfaces: dashboards, alerting with evidence, replanning tools, and audit trails
• Integration connectors across ERP/CMMS/WMS/historians/PLC layers (e.g., OPC-UA), including normalization + schema mapping

Responsibilities

• Ship production-quality features end-to-end across backend services and user-facing interfaces
• Build and improve real-time/streaming data architectures (Kafka/Confluent or similar)
• Partner closely with product and engineers to define scope, tradeoffs, and success metrics
• Operate with high autonomy (ambiguous problems, light specs) while maintaining a high quality bar
• Mentor others as needed while staying hands-on (preferred, not required)

What we’re looking for

• Zero-to-one delivery experience is required: you’ve shipped new products or major capabilities in an early-stage environment
• VC-backed startup experience strongly preferred (pre-seed through Series B ideal). Big tech logos alone are not considered a differentiator
• Agentic AI experience is non-negotiable: you’ve built and deployed AI systems/agents in production (orchestration, tool use, memory/state, human-in-the-loop)
• 3+ years of software engineering experience post-university (5+ preferred); internships can count toward total experience
• Full-stack engineering capability; comfortable across backend (Python/TypeScript) and frontend
• Experience integrating AI into enterprise environments (connectors, data systems, security/compliance constraints)
• Strong English communication skills (written and verbal); comfortable presenting and working cross-functionally (may interface with customers)

Nice to have / bonus

• Physical-world AI experience: video, IoT, sensor streams; industrial computer vision; robotics/autonomous systems
• Experience with real-time/streaming systems architectures
• Experience with ontology/knowledge graphs and structured context systems grounded in real asset/process hierarchies
• Background in high-performance, mission-driven teams (e.g., Palantir/Anduril-style environments)

Work arrangement

• Texas, then remote
• No immediate relocation required
• Visa sponsorship not available

Compensation

• Salary: $200K–$240K (OTE: $220K–$270K)
• Equity: 1.0%–1.5%
• Hiring: 3–4 candidates

Interview process

1. Initial discussion
2. Technical quiz (2–3 engineers)
3. Situational / culture interview
4. Conversations with multiple engineers

Why candidates should join

• Category-defining opportunity: build foundational software for industrial operations over the next 10–20 years
• Exceptional equity: employee equity pool meaningfully above industry standard
• Elite team: operators and builders with deep industry credibility across robotics, autonomy, and applied AI
• Real-world impact: make vast amounts of industrial sensor data useful in high-stakes environments