Solution Architect

## About Artikate Studio We're a deep-tech engineering firm that has been building software since 2016. We engineer AI-powered systems for defence institutions, government bodies, national-scale enterprises, and D2C brands — from a single AI module to entire mission-critical platforms. ## About the role We're hiring a Solution Architect to own the technical backbone of our most ambitious builds. You'll turn a brief into a buildable, secure, procurable system — the architecture, the integration design, the deployment topology, and the trade-offs — across our full range of work: AI/ML systems, defence command-and-control and computer vision, sovereign LLM driven systems, national-scale enterprise platforms, and digital commerce. This is hands-on and client-facing. You'll lead pre-kickoff technical reviews, scope hard problems under real security and integration constraints, and stay accountable for what ships. The people who scope our projects own them — that's you. ## What you'll architect - **End-to-end systems** across edge, on-premise, air-gapped, and cloud (AWS / GCP) — with the right boundaries between them. - **Computer vision & object detection** — multi-camera, multi-feed detection and inspection systems built for accuracy at scale. - **Integrated Command Systems** — fusing multi-sensor and multi-camera feeds into a single operational picture, with real-time event correlation, alerting, role-based operator consoles, and redundancy / failover / high-availability design. - **Sovereign LLM & RAG** — on-premise deployment of Llama-class models, GGUF quantisation, and RAG pipelines over private data corpora where no commercial cloud API is permitted. - **Air-gapped / LAN-only / zero-egress** deployments and classified or sensitive data handling. - **National-scale enterprise platforms** — high-concurrency systems, database sharding, telemetry/fleet intelligence, ERP and workflow automation, and analytics dashboards. - **Industrial & manufacturing vision** — machine-vision inspection and monitoring, PLC/CNC and ERP integration, and harsh-environment fitment where applicable. - **MLOps** — model registry, monitoring, and retraining pipelines; GPU-cluster architecture for on-prem and air-gapped inference. - **Security architecture** — security-first posture from day one, and the procurement-grade hardware specification / BOM where systems are physical. - Leading the **pre-kickoff brief, technical survey and scope definition**, and translating constraints into a written, shared design before code is written. ## Must-have experience - 4+ years architecting production systems across AI/ML and/or large-scale enterprise platforms — designs that actually shipped. - Strong computer vision: object detection and multi-feed systems (PyTorch, YOLO-family) and what survives real deployment. - LLM / RAG architecture: Llama-class models, RAG pipelines, LangChain, GGUF, Hugging Face. - Cloud architecture on AWS and/or GCP, plus Kubernetes, Terraform, Docker. - On-premise, air-gapped, and GPU-cluster deployment experience. - MLOps: model lifecycle, monitoring, retraining. - A genuine security mindset — air-gapped, zero-egress, and classified/sensitive data handling. - Backend fluency (FastAPI / Django / Node.js, PostgreSQL, Redis) and the ability to reason across the full stack. - Ability to run a client-facing technical scoping session and produce a clear written design and scope. ## Nice to have - Defence / government / command system experience — multi-sensor fusion, real-time situational awareness, HA/failover, and comfort with strict NDAs and classified environments. - Industrial / manufacturing systems — machine vision, PLC/CNC integration (OPC-UA, FANUC FOCAS, Modbus), and IATF-governed quality environments. - ERP / SAP integration. - Eligibility for, or willingness to undergo, security verification. ## How we work - **Senior-led delivery** — no hand-off to juniors after the pitch. - **2-week sprint velocity** — working software every fortnight, issues surface early. - **Defence-grade security baseline** — air-gapped and zero-egress are everyday work. - **Embedded partnership** — dedicated teams that treat the client's product as their own. - **IST-aligned** — no timezone friction, reachable during client business hours. ## Engagement details - **Location:** Gurugram, with travel to client sites as engagements require. - **Type:** Full-time, senior. - **Start:** Immediate.