Priyanka Arora — DevOps Engineer

AI Product Manager with ~10 years of experience building cloud-native and AI platforms, focused on shipping reliable, evaluation-driven GenAI systems from 0 → 1 → production → scale. I work at the intersection of product, ML, and infrastructure—treating AI not as features, but as end-to-end systems that must be measurable, controllable, and trustworthy. I’ve built and scaled GenAI platforms using RAG and agentic workflows, improving retrieval accuracy by 25%, reducing latency by 20%, and cutting model deployment time by 40%. My focus is on embedding AI into real decision workflows, driving 30%+ adoption by optimizing for usefulness, trust, and repeat usage—not just model capability. My approach is grounded in evaluation and iteration loops. I define “good” across multiple dimensions—accuracy, relevance, completeness, and safety—and build systems to continuously measure and improve them. I design for failure modes (hallucinations, low-retrieval confidence, ambiguity) and ensure systems degrade gracefully with clear fallback behaviors. • Build golden datasets and eval pipelines (LLM-as-judge + human review) • Instrument systems for continuous regression testing across prompts, models, and retrieval • Optimize trade-offs across latency, cost, and quality (model routing, caching, async flows) • Design guardrails: citation enforcement, confidence scoring, and safe fallbacks Technically hands-on, I work closely with engineering and data science on RAG architectures, retrieval optimization, prompt orchestration, and LLMOps pipelines (deployment, versioning, experimentation, monitoring). I’ve led cross-functional teams across product, DS, engineering, UX, and SRE—owning roadmap, prioritization, and execution in ambiguous, high-stakes environments. I’m particularly interested in building high-reliability AI systems where strong evaluation, fast iteration loops, and real-world constraints (latency, cost, safety) define product quality. If you're working on frontier GenAI systems, evaluation frameworks, or production-grade AI platforms— Always open to exchanging ideas on GenAI systems, evaluation strategies, and building reliable AI products at scale. Let's connect!

Stackforce AI infers this person is a SaaS Product Manager specializing in AI and cloud-native systems.

Location: Gurugram, Haryana, India

Experience: 10 yrs 6 mos