AI/ML Engineer

Role Overview
We are looking for an AI/ML Engineer to build, productionize, and optimize ML and Generative AI solutions that power intelligent, question‑driven analytics and workflow automation. You will design robust data/feature pipelines, implement LLM- and ML‑based services (including RAG and agentic patterns), and ship secure, explainable, and observable models into production—working closely with product, data, platform, and QA teams in an Agile environment.
Key Responsibilities
Model Development & Generative AI
Design, train, fine‑tune, and evaluate ML and LLM models for use cases such as intent classification, retrieval‑augmented generation (RAG), forecasting, recommendations, and anomaly detection.
Engineer prompts, system messages, and guardrails, and implement fallback strategies (e.g., safe completions, rules-based checks, defaults) to ensure reliability and usefulness.
Build agentic workflows that plan, call tools/APIs, reason over structured/unstructured data, and return explainable outputs.
Data, Features & EvaluationBuild reliable data/feature pipelines (batch & near‑real‑time) and maintain feature stores; ensure data quality, lineage, and reproducibility.
Establish offline/online evaluation: A/B tests, quality gates, bias/fairness checks, hallucination detection, and domain‑specific accuracy metrics.
Implement semantic/metadata alignment (business glossary, metric catalog, synonyms) so models interpret business questions consistently.
MLOps & Platform Engineering Own end‑to‑end model lifecycle: packaging, versioning, deployment, canary/A‑B rollout, drift detection, retraining, rollback, and cost/latency optimization.
Instrument observability (tracing, logging, metrics, LLM/ML telemetry) to monitor performance, safety, and usage; build dashboards and alerts.
Integrate with CI/CD pipelines (tests, security scans, infra-as-code), ensuring repeatable and compliant releases.
Security, Compliance & RBAC
Embed PII protection, RBAC inheritance, sample-size enforcement, peer-group rules, and audit trails in data/model services.
Contribute to risk assessments and responsible AI practices (explainability, human‑in‑the‑loop, model cards, usage policies).
Collaboration & Delivery