Laminar Hiring 5 Roles for Factory Floor AI: $114K–$180K for Senior Edge Platform

Why Laminar Is Hiring Now

Laminar builds ruggedized hardware and edge AI/ML solutions for the factory floor. Its 35+ employees span seven growing teams (mechanical, electrical, controls, firmware, ML, infrastructure, and product) and operate from the Boston area. Co-founders Annie and David Lu lead the team.

The company's public careers page lists a catalog of technical openings: Sensor Characterization Co-Op, (Jr) Machine Learning Developer, ML Infrastructure Engineer, Senior Software Engineer (Embedded Systems and Edge Platform), Product Engineer Full Stack, Junior Software Engineer Edge Platform, Head of Deployments, PLC Integration Engineer, Forward Deployed Engineer (CPG) hiring multiple, Junior R&D Engineer, and Senior Hardware Engineer. Every technical role touches the edge. No pure research track, no cloud-only backend, no data science role detached from deployment. The hiring plan is the strategy.

The company's stated values — "process driven," "egoless," "kaizen mentality," "self-starter," "endless empathy," "hungry" — read like a spec sheet for people who survive a 3 a.m. line-down call. Its manifesto: "Behind every small certainty is a system, and behind the system is a team that isn't firefighting. We hire the people who get the details right when nobody is watching, and we build the tools that let them." That sentence is the filter. The next section shows how Laminar applies it.

Inside the Screen: Five Stages, One Filter

Laminar runs a five-stage interview process built for engineers who ship hardened systems into environments where a kernel panic costs a production line hours of downtime.

Stage 1: a 15-to-20-minute phone screen by HR or a recruiter, a baseline check on communication and timeline. Stage 2: a 30-minute call with the hiring manager, focusing on technical depth and project ownership. Stage 3: the on-site, structured around a technical overview, a panel with the hiring manager and team members, and a skills exercise sent in advance. The exercise takes no more than an hour to prep. Laminar records the candidate's walkthrough and shares it with absent team members and founders before the final round. The recording becomes a durable artifact — evidence of how the candidate thinks under constraints, not just whether they produce a correct answer.

Stage 4: a Founder's Interview, in person with Annie or David Lu. Stage 5: two professional references (ideally one current, one former manager) and a background check.

The process is explicit about AI assistance: Laminar may use AI tools to review applications, analyze resumes, and flag inconsistencies. Those tools assist the recruitment team; they do not replace human judgment. Final hiring decisions remain human.

The role requirements show what the screen selects for. The Senior Software Engineer — Edge Platform opening lists three non-negotiable domains: Linux Systems Engineering (provisioning, networking, service management, storage, security, performance tuning on production edge fleets), Edge AI Deployment (running ML models and LLMs locally while accounting for compute, memory, latency, and hardware acceleration), and Industrial Integration (MES connectivity and protocols including Ethernet/IP, OPC UA, Profinet, Modbus, CC-Link). A candidate who has only trained models in the cloud will not clear the screen. The company's own framing makes the bar plain: "Models here must survive contact with physics. Hardware must survive years of continuous industrial operation. Software must integrate with machinery built decades ago. Everything we build ultimately has to work on a factory floor."

That philosophy extends to ownership. Laminar tells candidates: "We believe exceptional people should be given exceptional amounts of ownership. At Laminar, you will have the context to form your own view, the permission to challenge ours, and the resources to pursue the right answers, whatever technical or organizational boundaries stand in the way. There are few layers between identifying something important and changing it." The screen finds engineers who already operate that way: people who have chosen the hardest version of their discipline.

Why the Factory Floor Demands Edge AI

The factory floor does not tolerate latency. A beverage line filling thousands of cans per minute cannot wait for a cloud round-trip to adjust fill levels. A tire mold cycling in seconds cannot buffer video frames for remote inference. Cisco's industrial team puts it plainly: AI workloads here require deterministic performance, not variable latency; local execution, not dependency on external connectivity; strong isolation, not shared-risk architectures; scalable operations across multiple sites. When an AI model moves from simulation to physical actuation, physics takes over — vibration, temperature swings, electromagnetic interference, and the hard reality that a missed inference means scrap, downtime, or a safety incident.

Edge computing in manufacturing is projected to grow from $2.8 billion in 2024 to $25 billion by 2034 (24.6% CAGR). The hardware segment alone commanded $29.4 billion in 2025, driven by edge AI chipsets and GPU accelerators. Cloud-centric architectures hit a wall: transferring the sensor volume modern lines generate (on the order of 73 zettabytes by 2025) creates latency, bandwidth, and data-sovereignty problems that are "unscalable and unmanageable." The alternative is consolidated platforms where control logic (PLC/virtual PLC), visualization (HMI), SCADA monitoring, and AI workloads (vision, prediction, optimization) share ruggedized compute on the same substrate. Cisco's Unified Edge and similar platforms from Rockwell Automation, Siemens, and Schneider Electric are built for this convergence.

The use-case pull is measurable. Intel's survey of manufacturers found edge AI's most appealing applications cluster around real-time action: employee safety (58%), security (55%), smart inventory (51%), predictive quality control (51%), real-time equipment monitoring (49%), anomaly detection (49%). A tire manufacturer replaced manual inspection (90–95% accuracy, months of training) with edge computer vision that inspects 20,000 tires daily at 99.9% accuracy. A bottling plant deployed edge vision across nearly 1,000 supervision points to catch missing protective gear and unsafe movements without adding headcount. Predictive maintenance, the highest-adoption use case at $22.5 billion in 2025, cuts unplanned downtime by up to half. These are not dashboards. They are closed loops where inference triggers actuation in milliseconds.

Yet the talent to build and deploy these loops is scarce. Deloitte's 2025 survey found roughly 70% of manufacturers report moderate-to-significant hiring challenges across IT, OT, data science, engineering, application development, and cybersecurity. BCG warns that solution developers often lack shop-floor knowledge, producing technology misaligned with lean practices. Intel's Watts frames the gap directly: "You need people who understand the technology. You also need to make sure your IT people are working very closely with their OT manufacturing counterparts." The World Economic Forum projects a 4-million-worker deficit of AI-skilled industrial talent by 2027.

Laminar's openings reflect this exact intersection. The Forward Deployed Engineer (CPG) role, customer-facing, sits at the IT/OT boundary where models meet metal. The senior edge platform role and the ML Infrastructure Engineer positions target core systems engineering: ruggedized inference pipelines, sensor fusion, deterministic latency budgets that factory physics demand. Annie and David Lu have steered Laminar toward deployments where ML models optimize cleaning-in-place systems for food and beverage customers, yielding sustainability and profitability gains that only materialize when inference runs reliably on the line, not in a notebook. The hiring push signals a company moving from pilot to production scale, filling roles the market cannot easily staff.

Cloud AI vs. Edge AI: A Hiring Divide

The AI hiring boom is real. LinkedIn ranked "Artificial Intelligence Engineer" the fastest-growing job for 2025 and 2026; postings grew 163% year over year. Ravio measured an 88% surge in AI/ML hiring. But the boom is lopsided. Most roles sit at companies building on cloud APIs, fine-tuning LLMs, or shipping AI features inside SaaS products. Laminar hires for something else: ruggedized edge systems that run on factory floors where connectivity is spotty, latency is a hard constraint, and a model crash stops a production line.

The typical 2025 AI job description asks for prompt-engineering fluency, RAG pipeline experience, and comfort with OpenAI, Anthropic, or Gemini APIs. Layersrank flags three hiring mistakes that dominate the market: treating the role as a software engineer with AI on the résumé, over-weighting prompt engineering, and skipping cost and latency questions. Those mistakes make sense when the deployment target is a cloud VM with elastic compute. They make no sense when the target is an industrial gateway bolted to a CNC machine.

Laminar's open roles reflect that divergence. The careers page lists Senior Software Engineer Edge Platform, ML Infrastructure Engineer, Forward Deployed Engineer (CPG), and Junior Software Engineer Edge Platform, titles that signal systems work tied to physical deployment, not model specialization. The Forward Deployed Engineer role exists because customers need engineers who can integrate inference pipelines into PLC networks, tune models for thermal throttling on fanless hardware, and debug why a vision system drifts when coolant mist coats the lens. That profile — part firmware, part ML ops, part field applications — barely exists in the cloud-AI playbook.

Salary data underscores the split.

Levels.fyi puts AI-focused software engineers at $247k average compensation, 12.5% above non-AI peers at senior level. Specialized LLM developers average $209k. Those figures cluster around model development, API integration, and cloud-scale training, work that assumes GPUs are abundant and retries are cheap. Laminar's posted ranges reflect a different cost structure: smaller team, hardware BOM constraints, customers who pay for uptime, not token throughput.

The skill gap shows up in prior-role data. LinkedIn's 2026 report shows the most common feeders for AI engineers are Software Engineer, Data Scientist, and Full Stack Engineer, all cloud-native backgrounds. Laminar's team spans seven disciplines including mechanical, electrical, and controls engineering. Their careers page describes "interdisciplinary expertise to pioneering work in efficiency and sustainability." That's not a tagline; it's a hiring filter. A candidate who has only deployed containers to Kubernetes will not pass a screen that requires reasoning about Modbus TCP/IP, Profinet, OPC UA, EtherNet/IP, Linux provisioning on fanless edge gateways, and the thermal envelope of IP67-rated enclosures.

The industry's confusion about titles reinforces the contrast. "The instability is sharpest in the newest categories (LLM engineers, AI product engineers, foundation model roles) where professional consensus is still forming," notes the Newxel analysis. Laminar sidesteps that instability by not hiring for those categories at all. They hire for those two roles, plus PLC Integration Engineer, titles that signal: we need someone who can make the whole stack work in a dirty, loud, safety-critical environment.

That environment changes what "AI skills" means. The 2025 Stack Overflow survey found 84% of developers use or plan to use AI tools, but 46% distrust AI output accuracy and only 3% highly trust it. On a factory floor, distrust is a feature: you build watchdogs, fallback logic, and deterministic interlocks around the probabilistic model. Cloud-first AI hiring optimizes for model quality metrics (perplexity, F1, latency p99). Edge-first hiring optimizes for worst-case behavior when the model hallucinates and the conveyor keeps moving.

Sixty-six percent of business leaders say they would not hire someone without AI skills; 71% would take a less experienced candidate with AI skills over a more experienced one without. That pressure pushes everyone toward the cloud-AI template: add an LLM badge, ship a copilot, call it an AI product. Laminar's hiring push goes the other direction. They need engineers who can quantize a model to INT8, target edge silicon, and prove it meets a millisecond inference budget under thermal soak. The market is flooded with the former. The latter is what Laminar pays for.

Team Culture and Candidate Experience

Laminar's team page shows a 35-person group that tackles 100-mile charity rides, celebrates Oktoberfest after Drinktec, and holds rooftop lunches at Greentown Labs. The benefits package includes 100% individual health premium coverage, flexible PTO, $90/month transportation benefit, $65/month wellness benefit, 12 company-paid holidays, employer-matching 401(k), equity, and Greentown Labs membership. The company describes its culture as "work hard, play hard, and take care of yourself," a frame that matches the rigor of deploying hardened systems to global factory floors.

Working in frontier tech? Zero G Talent tracks the openings: see every open Laminar role, browse frontier tech jobs, the companies hiring, and the people building the field.