Skip to content

Abhinav Singarapu

AI Engineer • a************@gmail.com • +19******594 • drivetube.ai/•••••

Professional Summary

AI Engineer with 5+ years of experience designing and delivering enterprise AI platforms, multi-agent systems, Retrieval-Augmented Generation (RAG), and cloud-native data solutions. Specialized in intelligent retrieval, structured LLM orchestration, enterprise backend services, and scalable AI services using Python, FastAPI, PostgreSQL, and modern AI frameworks.

Technical Skills

Programming Languages: Python
Web Technologies: REST APIs
Frameworks and Libraries: scikit-learn,FastAPI
Databases: SQL,PostgreSQL,pgvector,Amazon Redshift,Azure Synapse Analytics
Cloud and DevOps: Amazon S3,Azure Data Factory,Databricks,ADLS,Docker
Data and Analytics: TF-IDF,BM25,Semantic Search,Embeddings,Cross-Encoder Reranking,AWS Glue,Apache Airflow,Apache Spark,Hadoop,Hive,ETL,Data Modeling,Pydantic
Tools and Methodologies: Git,GitHub
Skills: PySpark
Generative AI: Agentic AI,Retrieval-Augmented Generation,LangGraph,Gemini,MCP,LangSmith,Structured Tool Calling,Multi-Agent Orchestration

Work Experience

Axzora Resourcing
AI Engineer
Jan 2024 – Present
Built an enterprise AI assistant for knowledge retrieval, HR automation, and business workflow orchestration using agentic and RAG techniques.
Tech Stack: LangGraph, Gemini, LangSmith, BGE embeddings, BM25, Vector Search, PostgreSQL, pgvector, FastAPI, Pydantic, Docker
  • Designed and delivered an enterprise AI assistant enabling employees to retrieve organizational knowledge, automate HR operations, and execute business workflows via multi-agent collaboration using LangGraph, Gemini, FastAPI and PostgreSQL, improving automation and access to knowledge.
  • Implemented an end-to-end RAG pipeline with intelligent document parsing, semantic chunking, BGE embeddings, hybrid retrieval (BM25 + vector search) and cross-encoder reranking to increase retrieval relevance and produce grounded responses.
  • Orchestrated LangGraph Supervisor and Specialist agents to support dynamic routing, shared state management, conditional execution and retries, enabling robust multi-step reasoning and deterministic agent coordination in production.
  • Integrated Gemini Structured Tool Calling with Pydantic schemas to execute HR operations, validated data retrieval, calculations and memory services, enforcing schema-level validation and predictable tool execution.
  • Implemented PostgreSQL-backed persistence for user memory, business data, conversation history, execution state and agent logging to enable secure multi-user isolation, contextual personalization and auditability.
  • Deployed the platform with FastAPI and Docker and established production observability via LangSmith for routing validation, retrieval relevance checks, hallucination detection, centralized logging, health monitoring and resilient fallback strategies.
Capgemini
Data Engineer
Apr 2020 – Jul 2022
Built scalable ETL and data platforms for financial analytics, enabling enterprise reporting and downstream business intelligence at a consulting client.
Tech Stack: AWS Glue, Amazon S3, PySpark, Python, SQL, Amazon Redshift, Apache Airflow, Hive, Hadoop
  • Developed scalable ETL pipelines using AWS Glue, Amazon S3, PySpark, Python and SQL to ingest, transform and stage data for financial analytics and enterprise reporting.
  • Built distributed data transformation workflows implementing cleansing, joins, aggregations and partitioning across multi-stage S3 data lakes to improve processing throughput and reliability.
  • Designed analytical data models and loaded validated datasets into Amazon Redshift to enable high-performance analytics, BI queries and downstream reporting.
  • Implemented Apache Airflow DAGs to orchestrate ETL dependencies, scheduling, monitoring and automated recovery, reducing manual intervention and improving pipeline reliability.
  • Developed enterprise data quality frameworks performing schema validation, reconciliation, duplicate detection and exception handling to ensure end-to-end data consistency before loading to the warehouse.
  • Collaborated with business stakeholders and analysts to translate financial reporting requirements into BI-ready datasets and reliable data pipelines leveraging AWS Glue, Redshift and PySpark.

Education

Wright State University
Master of Science in Computer Science • 2024

Powered by Drivetube · Create your own profile at drivetube.ai