Building production-grade AI systems that are scalable, observable, and alignment-verified.

# developer_spec.yaml
identity:
  name: Harshada Javeri
  role: Senior Applied AI Engineer (7 YOE)
  specialization: Agent Systems, LLM Evals, & ML Platforms
telemetry:
  orchestration: [LangGraph, State-Machines, Multi-Agent Loops, Tool-calling]
  reliability:   [Continuous Evals, Drift Detection, Guardrails, Policy Adherence]
  infrastructure:  [Kubernetes, Docker, MLflow, AWS, CI/CD]
  observability:   [Prometheus, Grafana, Latency p99, Telemetry Metrics]

• Production Orchestration: Designing custom multi-agent loops, stateful graphs (LangGraph), and resilient tool-calling mechanisms.
• Deterministic Evals: Building framework-level safety gates, verification loops, and drift monitoring to tame stochastic model behavior.
• Observability: Scaling inference telemetry, anomaly detection, and regression testing for high-throughput AI workloads.

• MLOps: Automated model validation, pipeline orchestration, and experiment tracking.
• Infrastructure: Architecting containerized, cloud-native deployments with robust CI/CD quality gates.

Built a framework to trace, benchmark, and monitor complex agentic systems.

• Impact: Enabled automated regression detection, tracked state transitions, and enforced behavioral consistency across model updates.
• Tech: Python, LangGraph, CrewAI, MLflow, AgentOps

Designed a compliance-focused benchmarking framework for evaluating safety and policy adherence in agentic workflows.

• Impact: Automated tool-call validation, verified boundary conditions, and escalation pathways verification.

Engineered drift detection and latency tracking pipelines for enterprise ML deployments.

• Impact: Reduced post-deployment incidents by establishing automated quality gates and real-time model telemetry.

• AI & Agent Engineering: Python • LLMs • RAG • LangGraph • CrewAI • LLM Benchmarking • Safety Guardrails
• Core ML & Science: Scikit-Learn • NLP • Classification • Drift Detection • Feature Engineering
• Infra & MLOps: Docker • Kubernetes • MLflow • AWS • GitHub Actions • Jenkins
• Observability & Data: Prometheus • Grafana • SQL • PostgreSQL • Snowflake • REST APIs

• Email: harshada.javeri@gmail.com
• LinkedIn: linkedin.com/in/harshada-javeri-mle
• GitHub: github.com/harshada-javeri