Engineering Architect| GEN AI

Job description
GenAI Engineering Architect (Must have - AutoGen, CrewAI and WrenAI)

This role requires hands-on project expertise to implement an enterprise application built on top of SQL

and unstructured data (images,videos,logs etc.) using AutoGen, CrewAI, Azure OpenAI GPT-4 Turbo and

GPT-4V with PTUs. This is a hands-on architect role requiring both deep technical skills and the ability to

deliver complex AI applications end-to-end on large operational databases to render charts, tables and

other insights as completions from NLP-based prompts. Deep experience of Autogen and Azure AI Search

is a MUST. This is not a document retrieval, summarization or semantic search-based role.

Responsibilities:

1. Architectural Design:

• Collaborate with stakeholders to understand business requirements and translate them

into architectural blueprints.

• Design scalable, secure, and high-performance architecture for the Autogen-based LLM-

Integrated application.

• Define data models and schemas for integrating operational data from relational

databases into the application.

2. Implementation and Development:

• Lead the implementation efforts, ensuring adherence to architectural guidelines and best

practices.

• Develop robust APIs and interfaces for seamless communication between the application

and relational databases.

• Write efficient and maintainable code, following coding standards and version control

processes.

3. Integration and Testing:

• Integrate operational data from various relational databases into the application,

ensuring data consistency and integrity.

• Conduct thorough testing, including unit testing, integration testing, and performance

testing, to validate the functionality and scalability of the application.

• Troubleshoot and debug issues as they arise during the integration and testing phases.

4. Optimization and Performance Tuning:

• Identify performance bottlenecks and optimization opportunities within the application

architecture.

• Implement performance tuning strategies to improve the speed, reliability, and efficiency

of data retrieval and processing.

• Continuously monitor system performance and proactively address any degradation or

inefficiencies.

5. Documentation and Knowledge Sharing:

• Create comprehensive technical documentation, including architecture diagrams, API

specifications, and deployment procedures.

• Conduct knowledge sharing sessions to disseminate architectural knowledge and best

practices among team members.

• Provide guidance and mentorship to junior team members, fostering their professional

growth and development.Requirements:

• Must have : AutoGen Framework, CrewAI, WrenAI, SQL Agents, AG-Grid Flask / Django / Fast API

development expertise with least 2-3 project delivered as a lead developer / implementation

architect.

• Must have : Core Python – Iterators, Generators , OOP concepts, Python Shell (REPL) and Object

Relational Mapper, Data structure and Exception handling etc.

• Must have : AI Search, Vector Database creation for relational databases and unstructured data

• Must have : Azure app services expertise in terms of building and deploying AI apps using cloud

services.

• Must have : Deep expertise in Azure SQL, Azure Data Factory , Linked Services and Azure Synapse

etc.

• +-10 years of overall technology experience in core application development + AI project

architectural leadership of at least 3 years

• 5+ years’ experience leading development of AI application using Python backend frameworks

and multiple inferencing pipelines

• Rapid PoC/Prototyping skills and expertise in building and demonstrating application blueprints

without need a developer’s assistance.

• Deep, hands-on and architectural proficiency in Python, Ag-Grid and ReactJS

• Hands-on expertise of SharePoint indexes and data/file structures (Azure SQL)

• Good knowledge of Azure Form Recognizer for OCR of complex images, forms and other data

• Handson with implementing TaskWeaver, Autogen, Agentic Flows, Retrieval Augmented

Generation (RAG) and RLHF (Reinforcement Learning from Human Feedback)

• Designing and implementing vector databases on Azure cloud using Ai Search and Cosmos DB

vCore

• Sound project implementation level knowledge of Pinecone,FAISS,Weaviate or ChromaDB

• Deep expertise in Prompt Engineering using DsPy tools etc.

• Knowledge of NLP techniques like transformer networks, embeddings, intent recognition etc.

• Hands-on skills on Embedding and finetuning Azure OpenAI using MLOPS/LLMOPS pipelines.

• Strong communication, architectural sketching, and collaboration skills