Title: AI / ML Architect
About Birlasoft:
Birlasoft, a global leader at the forefront of Cloud, AI, and Digital technologies, seamlessly blends domain expertise with enterprise solutions. The company’s consultative and design-thinking approach empowers societies worldwide, enhancing the efficiency and productivity of businesses. As part of the multibillion-dollar diversified CKA Birla Group, Birlasoft with its 12,000+ professionals, is committed to continuing the Group’s 170-year heritage of building sustainable communities.
Job Title: AI ML Architect
Lactation:
Key Responsibilities:
Proficiency in Large Language Models (LLMs), Generative AI, Natural Language Processing (NLP), and Retrieval-Augmented Generation (RAG).
i. Knowledge of the leading LLMs. Fine-tuning models using Python libraries like Hugging Face Transformers or OCI's SDK to adapt them for domain-specific tasks
ii. NLP skills like sentiment analysis, entity extraction, and text classification
iii. RAG expertise – Understanding of Vectorization, embeddings. Working with OpenSearch, leveraging Oracle Database 23ai's vector capabilities. Experience with Prompt Engineering, Content Moderation Guardrails, and Handling Context Retention in Multi-Turn Interactions
iv. Advanced prompt engineering techniques, such as chain-of-thought prompting
v. Content Moderation: PII detection, toxicity scoring, custom filters
vi. Context Retention: expertise in managing session states, Redis Caching, multi – turn conversations.
vii. Tools Orchestration, Function Calling, Custom API Endpoints, SQL Tools for Database Queries, and Agent Networks for Multi-Stage Workflows
• Must have 13-15 years of relevant work experience.
• Drive Weekly, Monthly, Quarterly, Yearly Status Report calls with stakeholders.
• Excellent verbal, written and interpersonal communication skills with both technical and non-technical audiences.
• Ability to think tactically as well as strategically while coordinating cross functional teams in a matrixed environment.
• Developing standards and processes, encouraging (or enforcing where necessary) the use of those standards and processes.