Job Title: AI Engineer
Duration: 3 Months - Long Term
Location: Washington, DC 20433
Hybrid Onsite:
4 days per week from Day 1, with a full transition to 100% onsite anticipated soon.
Essential Job Functions:
Architect and Implement AI Solutions
Design and build RAG pipelines using Azure AI/Search and vector databases: chunking, embeddings, hybrid/semantic ranking, re-ranking, evaluation, and citation display.
Build enterprise conversational systems (multi-turn, retrieval-grounded) with prompt lifecycle management, guardrails, audit logging, and telemetry.
Support multiple LLMs and modalities: Azure OpenAI, Llama (Meta), Claude, etc.., and task-specific OSS models (vision, speech), with policy-driven model routing for performance, safety, and cost.
Integrate and Operate AI Infrastructure
Implement Model Context Protocol (MCP) servers integrating with project related areas.
Provide tool functions with RBAC scopes, schema versioning, rate limiting, request/response validation, and audit trails.
Deploy Azure AI Agent Service (AGA) patterns for agent registry/broker/governance with agent telemetry and policy enforcement.
Use Azure Batch for large-scale, parallel inferencing/vectorization jobs; leverage AWS EMR for distributed data/feature processing in AI pipelines.
Develop and Manage Data Pipelines
Build ingestion and enrichment for RAG connectors and ETL/ELT: document normalization, PII redaction, metadata enrichment, SLA/SLO monitoring, and lineage.
Operate large-scale vectorization with quality gates and drift monitoring.
Use Azure Data Factory (ADF) and Azure Databricks for orchestrated, scalable data processing; use AWS EMR for Hadoop/Spark workloads supporting AI features.
Build Agentic AI Solutions
Design secure tool-calling and multi-agent orchestration using Semantic Kernel, AutoGen, Microsoft Agent Framework, CrewAI, Agno, and LangChain or others.
Know how to apply agent governance and MCP-based controls across heterogeneous agents and runtimes (register, observe, govern, retire).
Model Evaluation and Optimization
Evaluate and fine-tune open-source and proprietary models; optimize for quality, latency, safety, and cost with A/B and offline eval suites.
Implement CI/CD with automated tests, security scans. Have knowledge on how to secure model workloads.
EEO: Mindlance is an Equal Opportunity Employer and does not discriminate in employment on the basis of – Minority/Gender/Disability/Religion/LGBTQI/Age/Veterans.