Staff Engineer(GENERATIVE-AI)

📋 Description

• • Own the end-to-end architecture of next-generation agentic AI products that will be deployed to Fortune-500 clients across North America, EMEA and APAC, influencing how millions of users interact with generative systems daily.
• • Design and implement robust, scalable Python micro-services that orchestrate multi-agent workflows, ensuring sub-second latency for real-time decision making while handling 10k+ concurrent API requests.
• • Lead the optimization of agentic engineering patterns—experiment with chain-of-thought, reflection, tool-use and multi-agent collaboration techniques to push accuracy benchmarks beyond 95 % on domain-specific tasks.
• • Fine-tune open-source and proprietary LLMs (Llama-3, GPT-4o, Claude, Gemini) using LoRA, QLoRA and RLHF; curate high-quality instruction datasets, run distributed training on GPU clusters and achieve measurable gains in retrieval-augmented generation (RAG) pipelines.
• • Architect seamless system integrations with enterprise ERP, CRM and data-lake ecosystems (SAP, Salesforce, Snowflake, Databricks) via REST, GraphQL and event-driven messaging, guaranteeing SOC-2-grade security and GDPR compliance.
• • Build resilient data-integration layers that ingest structured and unstructured data at terabyte scale, leveraging Spark, Kafka and Airflow to guarantee exactly-once semantics and 99.9 % uptime SLAs.
• • Collaborate with cross-functional squads of product managers, UX researchers and ML scientists to translate ambiguous business problems into crisp technical epics, delivering MVPs in 4-week sprints and full releases in 12-week quarters.
• • Champion code quality through rigorous test-driven development, peer reviews and automated CI/CD pipelines (GitHub Actions, ArgoCD) that deploy containerized services to AWS EKS and Azure AKS with zero-downtime blue-green releases.
• • Mentor a globally distributed team of 6–10 senior and mid-level engineers, running weekly architecture guilds, brown-bag sessions and pair-programming rotations to raise the bar on Pythonic craftsmanship and generative-AI best practices.
• • Contribute to Nagarro’s open-source AI accelerators, publish technical blogs and speak at global conferences (PyCon, ODSC, RE•WORK) to position the company as a thought leader in responsible and innovative generative AI.
• • Continuously monitor production telemetry (Prometheus, Grafana, LangSmith) to detect drift, hallucinations and latency anomalies; implement auto-remediation playbooks that cut incident MTTR to under 15 minutes.
• • Drive product-design thinking workshops with stakeholders to translate user pain points into delightful conversational interfaces, rapid-prototype in Figma and iterate based on A/B experimentation data.
• • Establish governance frameworks for prompt management, model versioning and ethical AI usage, ensuring every agentic system passes bias, fairness and safety audits before reaching production.
• • Evaluate emerging generative-AI research (tool-former architectures, mixture-of-experts, constitutional AI) and run quarterly spike projects to assess business applicability, filing internal patents where strategic value is proven.
• • Balance technical depth with business acumen—present quarterly OKR reviews to C-suite executives, quantify ROI of AI initiatives and secure multi-million-dollar follow-on budgets for next-gen capabilities.