Product Architect – AI-Powered Developer Tools (On-Prem)
Location: Tower 1, World Trade Center, Kharadi, Pune
Work Arrangement: Full-time, In-Office
Role Snapshot:
Be the architect who turns cutting-edge AI research into air-gapped, enterprise-grade developer tooling. You will research emerging AI-IDE innovations, spin up quick-and-dirty prototypes, and drive an 8-10-person team to ship those features at break-neck speed—the product mainly build to work within customer firewalls, customised LLM models and on-prem systems.
Key Responsibilities:
• On-Prem Architectural Ownership – design a fully self-hosted, distributed architecture that runs on VMs, bare-metal, or private Kubernetes/OpenShift clusters—data-plane to UI.
• Rapid Innovation Research – scout open-source / offline-capable AI techniques, prototype them in Python/TypeScript/Go, and distil viable features for the roadmap.
• Self-Hosted AI/ML Integration – deploy and fine-tune local LLMs (Llama-3, Mistral, Ollama, GPT-J, etc.), embeddings, and code-intel models with sub-second latency—no external calls.
• Secure Build & Packaging – craft hardened build pipelines, sign binaries, embed anti-reverse-engineering guards, and own SBOM & vulnerability closure.
• DevOps & CI/CD Leadership (Behind the Firewall) – standardise Docker images, private registries, and IaC (Ansible/Terraform/Helm) for on-prem clusters; enforce shift-left security.
• Technical Leadership – run design reviews, create ADRs, mentor and unblock an 8-10 dev team, keeping docs & diagrams one sprint ahead.
• Performance & Reliability – set SLOs, choose the right caches/queues, and tune until users feel performance is instant.
Ideal Candidate Traits:
• 8–12 years building large, scalable on-prem or hybrid dev-tool products; 3+ years as lead/principal architect.
• Fluent in at least two of Python, TypeScript/Node, Go, Rust, Java plus AI/ML frameworks.
• Proven track record of shipping self-hosted LLM/ML solutions without external API dependencies.
• Deep expertise in distributed systems, secure SDLC, Docker/K8s/OpenShift, and private IaCtooling.
• Thrives on turning research spikes into production-ready features—fast.
You'll be a good fit if you:
• Were a cracked builder before the Al wave.
• Have Al fully embedded in your workflow but didn't need it to be great.
• Know how to use tools like Cursor, Claude Code, Google CLI, Windsurf, Replit, and Lovable —and push them to the limit.
• Can build entirely functional prototypes within 1-2 hours using AI tools.
• Have built products people use (and love).
• Care more about who you work with than what you work on.
• Ship fast, iterate faster, and value momentum over perfection.
• See coding as craft — performance, polish, and UX matter to you.
• Are allergic to bureaucracy and energized by ownership.
Why Smart Analytica?
• Green-field canvas: define the DNA of a next-gen, on-prem AI coding platform.
• Hands-on freedom: founders expect you to lead and code.
• Tight-knit team: builders who care about craft, speed, and each other.
Ready to sketch the blueprint and ship it on-prem?
Email your résumé to hiring@smart-analytica.com with subject “Product Architect – ”
