Sr Software Engineer GenAI

Job Description

HYBRID – This role will be based out of The Leela Office located on the 4th Floor, Airport Road, Kodihalli, Bangalore- 560008. Our expectation at this time, is that you would work from our office on Tuesdays, Wednesdays, Thursdays with flexibility to work from home on Mondays and Fridays.

Mission: Own the intelligence and reliability of our AI stack — retrieval quality, evaluation, guardrails, and model operations. Ensure prototypes are accurate, safe, efficient, and production-ready. These roles are key to accelerating AI adoption across the organization by rapidly prototyping solutions and scaling those that deliver impact. Phase 1: Rapid prototyping for internal validation. Phase 2: Scaling validated prototypes into secure, production-ready, customer-facing applications.

Responsibilities

• Collaborate with cross-functional teams (Product, Engineering, CloudOps, BI, and internal stakeholders) to define use cases and deliver value.
• Design, build, and optimize RAG pipelines (document ingestion, embeddings, hybrid retrieval, re-ranking).
• Implement and orchestrate agentic AI workflows with tool integration.
• Build evaluation frameworks: golden sets, hallucination testing, red-teaming, prompt injection defense.
• Develop guardrails: policy checks, PII redaction, safety filters.
• Manage LLMOps: caching, model routing, cost/latency optimization.
• Support LLM fine-tuning (LoRA, PEFT) and dataset preparation.
• Collaborate with AI Product Engineer to deploy reliable prototypes.
• Harden PoCs into customer-facing systems with CI/CD, monitoring, rollback strategies.
• Implement security and compliance controls for SOC2, GDPR, HIPAA, and industry-specific standards.
  
  

• Ability to integrate with BI/data pipelines (Snowflake, Tableau, Informatica, or equivalent).
• RAG Expertise: Document ingestion, chunking, embeddings generation, retrieval metrics (recall@k, nDCG).
• Agentic AI Orchestration: LangChain, Autogen, CrewAI, function/tool calling.
• Evaluation & Guardrails: Golden sets, hallucination testing, red-teaming, prompt injection defense.
• Vector Databases: Pinecone, Weaviate, pgvector, FAISS.
• LLMOps & Optimization: Model routing (small→large), caching, latency/cost SLOs.
• Applied NLP: Text classification, summarization, sentiment, entity extraction.
• LLM Training: Instruction fine-tuning (LoRA/PEFT), dataset curation, data skewness and redundancy evaluations.
• Python Engineering: Strong in orchestration (pydantic/FastAPI), Pandas for data prep.
• Security/Compliance Awareness: RBAC, data privacy, audit logging.
• MLOps/Data Pipelines: Airflow, dbt, or lightweight ETL for ingest.
• Production Monitoring: Prometheus, Grafana, Datadog dashboards for AI workloads.
  
  

• Experience fine-tuning or training transformer models (LLMs, BERT, GPT variants).
• Knowledge of search/retrieval systems (Elasticsearch, OpenSearch, Vespa).
• Contributions to open-source AI/ML projects (LangChain, LlamaIndex, HuggingFace).
• Experience with benchmark datasets (SQuAD, GLUE, MS MARCO, or internal corpora).
• Hands-on with cloud AI services (AWS Bedrock, Azure OpenAI, GCP Vertex AI).
• Background in data security, compliance, and privacy standards (SOC2, GDPR, HIPAA).
• Strong math/ML foundation (optimization, statistics, linear algebra).