AI/ML Engineer

Job Details:
Overall 8 years’ experience with 5 years in AI development. PFB the technology skills required.
Core Language & Architecture
•                Python 3.11
•                Advanced type hints (PEP 484), static typing discipline
•                Async programming (asyncio, async/await, async generators)
•                aiohttp / httpx (async HTTP clients)
•                Pydantic v2 (BaseModel, validation, settings management)
•                Structured logging & tracing patterns
•                Redis (pub/sub, TTL, async clients)
•                REST API design & integration patterns
•                Retry/backoff strategies (Tenacity)
•                Concurrency patterns (parallel tool calls, task orchestration)

AI / LLM / Agent Systems
•                LangGraph (state machines, conditional edges, checkpointing)
•                LangChain 0.3.x (LLMChain, StructuredTool, retrievers, prompt templates)
•                ReAct-style agent architectures
•                Tool-based agent design (40 tool environments)
•                Azure OpenAI / OpenAI APIs (GPT-4o, deployment mgmt, rate limits, token budgeting)
•                Prompt engineering (few-shot, structured output, JSON mode)
•                PydanticOutputParser / structured LLM responses
•                Guardrails / PII redaction patterns
•                Memory abstractions for agents
•                Langfuse (trace instrumentation, evaluation, prompt management)
•                LLM fallback chains & error recovery
•                RAG prompt grounding strategies
•                LLM fine-tuning
•                Neural Network training & tuning
•                Traditional ML models (random forest, k-means clustering, linear regression, etc.)
•                MCP development and consumption

Retrieval, Search & RAG Engineering
•                Vector databases (Qdrant and/or Milvus)
•                HNSW indexing parameters
•                Filtering strategies
•                Embedding pipelines (OpenAI ada-002 or equivalent)
•                Batch embedding & re-indexing workflows
•                Hybrid retrieval (BM25 semantic)
•                Score fusion strategies
•                Cross-encoder reranking (BAAI/bge models)
•                FastAPI-based inference services
•                LangChain retriever abstractions
•                RAG evaluation metrics:
o               Faithfulness
o               Relevance
o               NDCG
o               MRR
•                Trace-level RAG evaluation (Langfuse)

Data Engineering & ETL
•                Prefect 2.x / 3.x
o               Flows, tasks, futures
o               Deployments (YAML)
o               Scheduling
•                ETL/ELT design
o               Schema evolution
o               Query optimization
•                OAuth authentication
•                Warehouse/schema management
•                PostgreSQL 16/17
o               psycopg 3.x
o               Connection pooling
o               SQLAlchemy 2.x (ORM asyncio)
o               Alembic migrations
o               Advanced SQL
o               Multi-table JOINs
o               CTEs
o               Window functions
•                Timezone conversion
•                Pandas 2.x (complex multi-stage transformations)
•                PyArrow / columnar formats
•                Azure Blob Storage (azure-storage-blob)
•                Document ingestion/parsing:
o               Docling
o               Unstructured
o               python-docx
o               python-pptx
 
DevOps & Platform
•                Docker
•                Linux fundamentals
 
 Nice-to-Haves
•                Ray (distributed execution)
•                Columnar performance tuning
•                Network operations domain knowledge
•                NOC / alarm correlation familiarity

 API & Enterprise Integrations
•                OAuth 2.0 (client credentials flow, token lifecycle)
•                MSAL (browser service principal flows)
•                Microsoft Graph API
•                SharePoint
•                Outlook
•                Planner
•                OneDrive
•                Pagination
•                App permissions
•                ServiceNow REST API
•                Table API
•                Incident/change mgmt
•                Bulk operations
•                Splunk SDK
•                Saved searches
•                Async queries
•                Log analysis
•                Azure AD app registrations
•                IPAM / OTNA integrations (nice-to-have domain exposure)