Lead Generative AI EngineerPylon Management Consulting • Delhi, IN
Lead Generative AI Engineer
Pylon Management Consulting • Delhi, IN
29 days ago
Job description
- 5+ years building cloud applications with at least good years on GCP.
- Delivering GenAI solutions in production (chat, copilots, summarization, extraction, or agentic workflows).
- Strong with Vertex AI : training/tuning (custom or LoRA/adapter), endpoints, batch/online prediction, embeddings and Vector Search, Workbench/Notebooks.
- Proficiency in Python (FastAPI/Flask), or Node.js; comfortable with REST/gRPC and event-driven designs.
- IaC with Terraform (modules, workspaces, state management) and secured CI/CD (Cloud Build/Cloud Deploy or GitHub Actions).
- Containers and orchestration : Docker and GKE or Cloud Run; understanding of service meshes, autoscaling, and rollout strategies.
- Data stack : BigQuery, Cloud Storage, Dataflow/Beam basics; feature/embedding pipelines.
- Security/compliance fundamentals : IAM, VPC Service Controls, CMEK/KMS, artifact signing, network policies, and secrets.
- Solid understanding of RAG, vector databases, prompt engineering, evaluation metrics, and guardrails.
Preferred Qualifications :
- Experience with multi-agent/agentic frameworks and tool-calling; LangChain/LangGraph or equivalent.
- Familiarity with governance and safety tooling (content moderation, toxicity filters, jailbreak/PII detection).
- Observability : OpenTelemetry, SLO design, chaos/resiliency testing.
- MLOps tools : MLflow/Vertex ML Metadata, experiment tracking, model registry.
- Performance tuning for LLM inference: batching, quantization, caching, model distillation, cost controls.
- Additional cloud exposure (AWS Bedrock or Azure OpenAI) and hybrid patterns.
- Prior work in regulated environments (HIPAA, SOC 2, FedRAMP) a plus.
Nice-to-Have Tech :
- Frameworks : LangChain/LangGraph, Ray, Beam.
- Datastores : Bigtable, Firestore, Redis, Elasticsearch, AlloyDB.
- Frontend integration for AI UX (React) and conversation memory patterns.
- Testing/eval : prompt/unit tests, golden datasets, offline/online evals, AB testing.
What Youll Work On (Examples) :
- A Vertex AIpowered RAG service using BigQuery + Vertex Vector Search, served on Cloud Run with Terraform-managed infra and Cloud Build CI/CD.
- An agentic workflow that orchestrates tools (REST/GraphQL/gRPC) to complete multi-step tasks with audit trails and human-in-the-loop review.
- A secured inference platform with per-tenant isolation, CMEK, egress controls, and cost dashboards. (ref:hirist.tech)
Job Description Qualifications :
- 5+ years building cloud applications with at least good years on GCP.
- Delivering GenAI solutions in production (chat, copilots, summarization, extraction, or agentic workflows).
- Strong with Vertex AI : training/tuning (custom or LoRA/adapter), endpoints, batch/online prediction, embeddings and Vector Search, Workbench/Notebooks.
- Proficiency in Python (FastAPI/Flask), or Node.js; comfortable with REST/gRPC and event-driven designs.
- IaC with Terraform (modules, workspaces, state management) and secured CI/CD (Cloud Build/Cloud Deploy or GitHub Actions).
- Containers and orchestration : Docker and GKE or Cloud Run; understanding of service meshes, autoscaling, and rollout strategies.
- Data stack : BigQuery, Cloud Storage, Dataflow/Beam basics; feature/embedding pipelines.
- Security/compliance fundamentals : IAM, VPC Service Controls, CMEK/KMS, artifact signing, network policies, and secrets.
- Solid understanding of RAG, vector databases, prompt engineering, evaluation metrics, and guardrails.
Preferred Qualifications :
- Experience with multi-agent/agentic frameworks and tool-calling; LangChain/LangGraph or equivalent.
- Familiarity with governance and safety tooling (content moderation, toxicity filters, jailbreak/PII detection).
- Observability : OpenTelemetry, SLO design, chaos/resiliency testing.
- MLOps tools : MLflow/Vertex ML Metadata, experiment tracking, model registry.
- Performance tuning for LLM inference: batching, quantization, caching, model distillation, cost controls.
- Additional cloud exposure (AWS Bedrock or Azure OpenAI) and hybrid patterns.
- Prior work in regulated environments (HIPAA, SOC 2, FedRAMP) a plus.
Nice-to-Have Tech :
- Frameworks : LangChain/LangGraph, Ray, Beam.
- Datastores : Bigtable, Firestore, Redis, Elasticsearch, AlloyDB.
- Frontend integration for AI UX (React) and conversation memory patterns.
- Testing/eval : prompt/unit tests, golden datasets, offline/online evals, AB testing.
What Youll Work On (Examples) :
- A Vertex AIpowered RAG service using BigQuery + Vertex Vector Search, served on Cloud Run with Terraform-managed infra and Cloud Build CI/CD.
- An agentic workflow that orchestrates tools (REST/GraphQL/gRPC) to complete multi-step tasks with audit trails and human-in-the-loop review.
- A secured inference platform with per-tenant isolation, CMEK, egress controls, and cost dashboards. (ref:hirist.tech)
Create a job alert for this search
Lead Generative AI Engineer • Delhi, IN