APPIT Software Inc.
Website:
appitsoftware.com
Job details:
FREELANCE
Job Title: Automation Workflow / AI Engineer
Timing: General IST (2:00 PM – 11:00 PM)
Experience: 8+ years
Role Summary
We are seeking a Workflow / AI Engineer to design, build, and operate AI-powered workflows with strong ownership over prompts, agents, schema validation, cost guardrails, and reference workflows. This role focuses on delivering reliable, observable, and secure AI workflows for production use.
Key Responsibilities
- Build AI-assisted workflow steps including extraction, classification, summarization, decision support, and next-action recommendations.
- Design and operate workflows using n8n with reusable sub-workflows and queue mode.
- Enforce “glass-box” controls for AI usage:
- Deterministic validation gates before and after AI calls
- JSON schema validation for structured outputs
- Allowlisted tools and actions only
- Human approval for high-risk actions
- Implement prompt versioning, model/version pinning, and rollback mechanisms.
- Build evaluation and regression testing frameworks using golden datasets, offline evaluations, and prompt tests.
- Implement drift monitoring and failure triage for AI workflows.
- Design and implement RAG (Retrieval-Augmented Generation) when required, with citations and grounding.
- Operate AI workflows in production with full traceability, observability, and cost controls.
- Ensure PII redaction, secure secrets management, and audit logging.
- Build and maintain CI/CD pipelines and infrastructure-as-code for workflow deployments.
- Document reference workflows, prompts, and operational standards.
Core Technical Skills (Must-Have)
Workflow Orchestration:
- n8n with reusable sub-workflows and queue execution
LLMs & AI:
- Strong experience with at least one LLM provider (OpenAI / Claude / Gemini)
- Structured output (JSON schema) and tool/function calling
Languages:
- Node.js, TypeScript / JavaScript
- Python (nice-to-have for evaluation tooling)
Vector / Retrieval (Preferred):
- Pinecone / Weaviate / Milvus / pgvector / Vertex AI Vector Search
Data & Storage:
- PostgreSQL, object storage (S3), Redis
Observability:
- OpenTelemetry with Grafana or Datadog
- Model-call metrics (latency, token usage, cost)
Security:
- PII redaction
- Secrets management (Vault / AWS Secrets Manager)
- Audit logs
CI/CD & Platform:
- GitHub Actions / GitLab CI / Jenkins
- Terraform, Helm
Required Qualifications:
- 4+ years of experience in workflow/integration engineering with hands-on delivery of LLM-based solutions.
- Strong experience with at least one LLM provider (OpenAI, Claude, or Gemini).
- Proven experience with structured outputs (JSON schema), tool/function calling, and guardrails.
- Experience building CI/CD pipelines, automated testing, observability, and secure production operations.
Behavioural Expectations:
- Strong ownership of AI workflows end-to-end
- Focus on reliability, safety, and cost efficiency.
- Clear documentation and communication
- Ability to troubleshoot production issues independently.
Click on Apply to know more.