Post Job Free
Sign in

Senior Generative AI Engineer (RAG/Agentic)

Location:
Plano, TX
Posted:
November 16, 2026

Contact this candidate

Resume:

LAKSHMI

[469-***-****] [***************@*****.***] [LinkedIn]

PROFESSIONAL SUMMARY

Senior Generative AI Engineer and Machine Learning Engineer with 10+ years of progressive experience designing, building, and deploying production-grade AI systems, conversational AI applications, agentic AI workflows, and enterprise data platforms across Healthcare, Financial Services, Retail, and Technology sectors. Deep expertise in Generative AI, Retrieval-Augmented Generation

(RAG), agentic AI architectures, prompt engineering, multi-modal AI, vector databases, semantic retrieval, and LLM evaluation and observability. Proven track record integrating large language models, including Google Gemini, GPT-4/4o, Claude, LLaMA, and Mistral, with enterprise data sources, APIs, and business systems using GCP Vertex AI, AWS Bedrock, and Azure OpenAI. Strong experience implementing AI guardrails, security controls, governance, monitoring, and human-in-the-loop workflows while collaborating with cross-functional stakeholders to deliver scalable enterprise AI solutions. TECHNICAL SKILLS

• Generative AI & LLMs: OpenAI (GPT-4, GPT-4o, GPT-3.5, Embedding’s), Anthropic Claude, Google Gemini, Meta Llama 3/3.1, Mistral, Falcon, Cohere Command R+, Amazon Titan, AWS Bedrock, Azure OpenAI, GCP Vertex AI.

• Google & Enterprise GenAI: Google Gemini, GCP Vertex AI, BigQuery, Generative AI, Conversational AI, Agentic AI, RAG, Function Calling, Tool Use, Enterprise AI Integration

• RAG & Knowledge Systems: RAG (Naïve, Advanced, Modular, Corrective, Self-RAG, HyDE, RAPTOR), Hybrid Search (BM25 + Vector), Re-ranking (Cohere R]erank, Cross-Encoder), Contextual Compression, Parent-Document Retrieval, Multi-Query Retrieval, Query Decomposition, Metadata Filtering, Document Chunking, Citation & Grounding.

• Agentic AI & Orchestration: Lang Chain (LCEL, Lang Graph, Agents, Tools, Memory), Llama Index, Auto Gen, Crew AI, Semantic Kernel, OpenAI Assistants API, Lang Smith, Haystack.

• LLM Fine-Tuning & Optimization: Hugging Face Transformers, PEFT (LoRA, QLoRA, Prompt/Prefix Tuning), TRL (SFT, PPO, DPO), bitsandbytes, Un sloth, GPTQ, GGUF/llama.cpp, vLLM, Text Generation Inference (TGI), Ollama, Flash Attention 2.

• Prompt Engineering & Evaluation: System Prompts, Zero/Few-Shot, Chain-of-Thought (CoT), React, Function Calling, JSON Mode, Prompt Chaining, Prompt Compression, Adversarial Testing, Constitutional AI, RAGAS, TruLens, Deep Eval, Prompt foo, Lang Smith, Arize AI, ML flow, Weights & Biases (W&B), Human Evaluation, A/B Testing.

• Vector Databases & Search: Pinecone, Qdrant, we aviate, Chroma, FAISS, pgvector, Amazon OpenSearch, Azure AI Search, Elastic search, Milvus.

• Machine Learning & NLP: Python, Scikit-learn, Tensor Flow, Keras, PyTorch, PyTorch Lightning, Hugging Face (Datasets, Accelerate, Evaluate), XGBoost, Light GBM, Cat Boost, BERT, RoBERTa, De BERT, Sentence Transformers, spacy, NLTK, Feature Engineering, Hyper parameter Tuning (Optuna, Hyperopt), SHAP, LIME, Time Series (Prophet, ARIMA, LSTM).

• Data Engineering: Python (Pandas, NumPy, SQLAlchemy, boto3, PySpark, Pytest), Data bricks (Unity Catalog, Delta Lake, Delta Live Tables, ML flow, Workflows), Apache Kafka, Apache Airflow, dbt, Great Expectations, Snowflake.

• Cloud, MLOps & DevOps: AWS (S3, Glue, EMR, Redshift, Lambda, Step Functions, Kinesis, Sage Maker, Bedrock, Athena, Dynamo DB, Cloud Watch, Secrets Manager, CDK), Azure (ADLS Gen2, Data Factory, Databricks, Synapse, Azure ML, Azure OpenAI, Event Hubs, Key Vault), GCP (BigQuery, Dataflow, Cloud Storage, Vertex AI, Cloud Composer, Pub/Sub), Docker, Kubernetes (EKS, AKS), Helm, Terraform, GitHub Actions, Azure DevOps, Jenkins, FastAPI, Flask, Streamlit, Gradio.

• Multimodal AI: OpenAI Vision, Claude Vision, AWS Textract, Amazon Recognition, Azure Document Intelligence, Whisper, DALL·E 3, Stable Diffusion, Lang Chain Multimodal.

• Databases & SQL: PostgreSQL, MySQL, SQL Server (T-SQL), Oracle, MongoDB, Dynamo DB, Redis, Elastic search, Snowflake SQL, Spark SQL, BigQuery SQL.

• Security & Responsible AI: Guardrails AI, NVIDIA NeMo Guardrails, Prompt Injection Prevention, PII Protection, Azure Responsible AI, HIPAA, SOX, GDPR, AI Governance.

• Tools & Visualization: Git, GitHub, GitLab, JIRA, Confluence, Postman, Swagger/Open API, Jupyter Notebook, VS Code, PyCharm, Lang Chain Hub, Hugging Face Hub, Power BI, Tableau.

• Semantic Data & Metadata: Semantic Retrieval, Semantic Search, Metadata Filtering, Data Modeling, Metadata Management, Enterprise Data Integration, Snowflake, dbt, BigQuery PROJECT EXPERIENCE

Client: Galen college of Nursing Location: Louisville, KY Role: Senior Generative AI Engineer Duration: July 2024 – July 2026 Project Overview:

Served as Senior Generative AI Engineer for a major US health system’s Clinical AI division, designing and deploying HIPAA- compliant, production-grade applications on AWS. Built a Clinical Knowledge Co-Pilot conversational assistant, an AI documentation tool for SOAP notes and prior authorization letters, a multi-agent Clinical Trial Matching engine, and an LLM-driven Patient Risk Stratification service, all validated via clinical safety evaluation frameworks. Key Contributions & Responsibilities:

• Architected the Clinical Knowledge Co-Pilot, a production RAG system for clinicians using Llama Index for medical section- aware chunking, Amazon Bedrock Titan Embedding’s V2, and Qdrant on AWS EKS with hybrid dense + BM25 retrieval and RRF re-ranking.

• Designed and developed enterprise conversational AI and agentic workflows integrating LLMs with structured and unstructured data sources, APIs, vector databases, and business systems using LangChain, LangGraph, LlamaIndex, and cloud-native AI services.

• Grounded LLM generation with Claude Sonnet 3.5 via AWS Bedrock, enforcing source citations and uncertainty handling in system prompts.

• Implemented HIPAA PHI de-identification guardrails using Guardrails AI and a fine-tuned BioBERT NER model, redacting PHI before Bedrock calls and re-injecting tokens for authorized display.

• Built the Clinical Documentation Assistant with Lang Chain LCEL: AWS Textract for scanned forms, HL7 FHIR R4 data from Epic, SOAP note generation, insurer PA letter output via tool use, and Claude Haiku quality scoring; cut documentation time from 18 to 4 minutes per encounter.

• Developed the Clinical Trial Matching Agent using Lang Graph multi-agent workflow: profile extraction, ClinicalTrials.gov search, LLM eligibility evaluation, ranking, and human-in-the-loop coordinator review with feedback re-ranking.

• Fine-tuned BioLinkBERT-large for Clinical NER with Hugging Face PEFT/LoRA on AWS SageMaker, achieving macro F1 0.91 and deploying as a real-time endpoint for query expansion.

• Designed RAPTOR hierarchical retrieval for long clinical guidelines: recursive clustering, UMAP visualization, abstractive summaries with Claude Haiku, and level-aware retrieval via Lang Chain Router Chain; improved RAGAS relevancy from 0.78 to 0.87.

• Implemented semantic chunking using sentence-transformers/all-MiniLM-L6-v2 cosine similarity to preserve clinical context boundaries, raising RAGAS Context Precision from 0.72 to 0.83.

• Built a comprehensive evaluation framework: RAGAS CI/CD gate, Lang Smith tracing, weekly clinician human review with rubric scoring, and monthly Prompt foo red teaming for safety and injection testing.

• Created a production FastAPI service with streaming /chat via Server-Sent Events, async /documents/ingest via SQS + Lambda, session storage in Dynamo DB, and JWT RBAC via AWS Cognito with Qdrant payload filters.

• Deployed services on AWS Fargate (ECS) with multi-stage Docker, auto-scaling on ALB metrics, health checks, and CodeDeploy blue-green releases with rollback.

• Orchestrated multi-step GenAI pipelines with AWS Step Functions for documentation and trial matching, including retries, parallel eligibility checks, and conditional human review.

• Applied model tie ring and cost optimization in AWS Bedrock: routing to Claude Haiku or Sonnet 3.5, semantic caching in Redis Elastic Cache, and budget alerts via AWS Budgets + SNS; reduced spend 41% while supporting 3x growth.

• Provisioned infrastructure as code with Terraform: ECS, EKS, RDS PostgreSQL/pgvector, Dynamo DB, Lambda, Secrets Manager, and least-privilege IAM; delivered via GitHub Actions CI/CD with RAGAS gates and sub-12-minute deployments.

• Built multi-modal processing using AWS Textract for forms/tables and Claude Vision via Bedrock for lab report images with charts and abnormal flags.

• Collaborated with clinical informatics, compliance, legal, and product teams to define AI Clinical Use Policy, IRB data agreements, and present safety frameworks to CISO, CMO, and Board Technology Committee.

• Mentored 3 AI and 2 ML engineers through 1:1 coaching, bi-weekly sessions on RAG, Lang Graph, RAGAS, and prompt safety, and established team standards for logging, evaluation, and PR reviews. Tech Stack: Python, AWS Bedrock (Claude Sonnet 3.5/3, Haiku, Titan Embedding’s V2), Lang Chain (LCEL, Lang Graph — agent state machines), Llama Index (multi-document RAG), Qdrant (vector store — self-hosted on EKS), pgvector (PostgreSQL), FastAPI

(deployed on AWS ECS Fargate), OpenAI GPT-4o (secondary LLM — Azure OpenAI endpoint), Hugging Face (Bio BERT, Clinical BERT

— NER and classification), AWS (S3, Lambda, Step Functions, Event Bridge, SQS, Sage Maker, Dynamo DB, RDS PostgreSQL, ECS/Fargate, ECR, Cloud Watch, Secrets Manager, Lake Formation), RAGAS, Lang Smith, Guardrails AI, Weights & Biases, Docker, Terraform, GitHub Actions, HIPAA, HL7 FHIR R4, 21 CFR Part 11 awareness Client: Goldman Sachs Location: Richardson TX

Role: Generative AI Engineer Duration: September 2022 – June 2024 Project Overview:

Served as a Generative AI Engineer within a regional bank’s Digital Innovation division, designing and building Azure-deployed, SOX, and PCI-DSS compliant financial intelligence systems. Developed an LLM-powered Financial Intelligence Platform for relationship managers, a Regulatory Compliance AI Assistant for automated Q&A and policy gap analysis (BSA/AML, FDIC), a natural language- driven Fraud Investigation Copilot, and an automated customer onboarding document processing pipeline. Key Contributions & Responsibilities:

• Designed enterprise agentic AI architectures integrating conversational AI agents with structured data platforms, vector search, APIs, and business systems, implementing secure tool calling, role-based access, observability, and governance controls for production AI workloads.

• Architected the AI Financial Intelligence Platform on Azure with multi-source RAG: ingested 15,000 regulatory filings and reports using Azure AI Document Intelligence, embedded with Azure OpenAI text-embedding-3-large, and indexed in Azure AI Search with hybrid vector + BM25 + semantic ranker.

• Built a Lang Graph Financial Intelligence Agent with State Graph routing: query classification knowledge base retriever or structured data queries or hybrid response synthesis via GPT-4o.

• Implemented a SQLAlchemy-based Snowflake query tool within Lang Chain to enable natural language access to loan, deposit, NIM, and capital metrics with parameterized queries and table whitelisting.

• Developed the Regulatory Compliance AI Assistant using Llama Index Router Query Engine to route questions to federal, internal, or hybrid engines; added a Lang Chain gap-analysis chain producing structured JSON assessments with remediation recommendations.

• Created an automated regulatory monitoring pipeline with Azure Functions triggered by FDIC and Federal Register RSS feeds, parsing new releases and upsetting to Azure AI Search with Teams alerts for relevant changes.

• Built the Fraud Investigation Copilot as a React Lang Graph agent with custom tools: Transaction Query Tool on PostgreSQL, Account Network Tool on Redis graph, Case History Tool via Azure AI Search, Alert Look up Tool for AML data, and Fraud Report Generator Tool; reduced investigation time from 2.5 hours to 28 minutes.

• Engineered GPT-4o Vision KYC pipeline with Azure AI Document Intelligence: ID and proof-of-address extraction, cross- verification of fields, consistency checks across documents, and OFAC screening integration to block high-risk onboarding.

• Applied advanced prompt strategies: CoT prompting for regulatory interpretation, few-shot examples for ratio analysis, JSON structured output with Pydantic validation, and LLMLingua prompt compression cutting tokens by 57%.

• Fine-tuned a sentence-transformers bi-encoder with PEFT/LoRA on Azure ML for financial semantic similarity, improving Context Recall from 0.76 to 0.89 on the compliance QA benchmark.

• Configured Azure AI Search hybrid search at 4.2M vectors: HNSW indexing, RRF fusion of vector and BM25 scores, Semantic Ranking re-ranking, and metadata filters by jurisdiction and effective date.

• Delivered data engineering support via Azure Data Factory for regulatory ingestion, dbt models on Snowflake for the financial analytics Gold layer, and Great Expectations data quality gates.

• Developed FastAPI back ends for chat, compliance gap analysis, and fraud investigation with SSE streaming, a sync Celery jobs, and Azure AD OAuth 2.0 role-based access; deployed on AKS with Helm and Horizontal Pod Auto scaler.

• Implemented SOX compliance controls: immutable audit logging to Azure Blob Storage WORM, Lang Smith tracing for every interaction, and explain ability with source citations in all AI outputs.

• Established evaluation and MLOps: RAGAS and Lang Smith automated evaluations in Azure DevOps CI/CD with deployment gates, and W&B tracking for fine-tuning experiments and model registry.

• Led Terraform IaC for AKS, Azure AI Search, Azure OpenAI, Key Vault, and Event Hubs; managed secrets via Default Azure Credential and zero hardcoded credentials.

• Collaborated with product, compliance, legal, risk, and security teams to define AI governance tiers, explain ability standards, and complete the bank's Model Risk Management validation for production deployment. Tech Stack: Python, Azure OpenAI (GPT-4, GPT-4o, text-embedding-ada-002, text-embedding-3-large), Lang Chain (Agents, Tools, LCEL, Lang Graph), Llama Index (Router Query Engine, Sub Question Query Engine), Azure AI Search (vector + hybrid search — semantic ranking), Qdrant (self-hosted), FastAPI (AKS deployed), Azure (ADLS Gen2, ADF, Databricks, Azure ML, Azure OpenAI, Azure AI Search, Azure Functions, Event Hubs, Key Vault, Azure Monitor, AKS, ACR, Azure DevOps), Snowflake, dbt, Apache Kafka (Event Hubs Kafka interface), Great Expectations, Python (Pandas, SQLAlchemy, Psycopg2), PostgreSQL, Redis (Azure Cache), Lang Smith, RAGAS, Guardrails AI, Weights & Biases, Terraform, Docker, Kubernetes (AKS), GitHub Actions, Azure DevOps, SOX compliance, PCI- DSS awareness

Client: Walmart Location: Rogers, AR

Role: Machine Learning Engineer Duration: March 2020 – August 2022 Project Overview:

Served as a Machine Learning Engineer within a major US Omni channel retailer’s Data Science team, owning the end-to-end ML lifecycle from feature engineering to Amazon SageMaker deployment and monitoring. Designed and deployed production systems including a real-time product recommendation engine, a multi-horizon demand forecasting system (Light GBM/LSTM ensemble), an XGBoost-based customer churn propensity model, and a BERT-based bi-encoder semantic search system for product discovery. Key Contributions & Responsibilities:

• Designed and productionized a real-time product recommendation engine for 15M+ MAU: ALS Matrix Factorization on EMR for collaborative filtering, Sentence-BERT embedding for semantic similarity, and Light GBM Lambda MART blending with business rules.

• Served recommendations via SageMaker real-time endpoint with Redis ALS vectors and FAISS semantic search, achieving p99 latency <80ms at 5,000 RPS.

• Built a multi-horizon demand forecasting system for 500K+ SKU-store combinations: PySpark feature engineering on EMR with 80+ features including lags, seasonality, promotion, and NOAA weather.

• Trained Light GBM with Optuna and a PyTorch Bi-LSTM on Sage Maker; ensemble at 0.65/0.35 weights to reach 6.8% SMAPE on 7-day forecasts.

• Automated weekly retraining and deployment with Sage Maker Pipelines, Event Bridge, and conditional model registration based on SMAPE improvement.

• Developed a 90-day churn propensity model for 3.2M loyalty members using XGBoost with class imbalance tuning; achieved AUC 0.84 and Precision@10% 0.67.

• Delivered SHAP explain ability for churn and forecasting models, surfacing top drivers in dashboards for marketing and planning teams.

• Stored and served real-time features in SageMaker Feature Store with online store for inference and offline S3 Parquet for training.

• Built a BERT-based semantic product search: fine-tuned msmarco-distilbert-base-v4 on 150K query-product pairs, batch- generated embedding with Sage Maker Batch Transform, and indexed in FAISS IVF-PQ.

• Implemented hybrid retrieval combining FAISS semantic similarity with Elastic search BM25 using RRF; A/B testing cut zero- result rate from 8.3% to 1.2% and lifted add-to-cart by 11%.

• Created FastAPI inference services for recommendations, churn risk, and search, deployed on EKS with KEDA auto scaling from 2 to 18 pods during peak.

• Engineered real-time data pipelines: Kinesis Data Streams + Lambda for clickstream to Dynamo DB, and Kafka MSK for inventory events updating features in Feature Store within 5 minutes.

• Orchestrated batch workflows with Airflow DAGs for churn training and daily recommendation cache warm-ups.

• Implemented MLOps with Sage Maker Model Monitor for data drift using PSI, Model Registry for lineage, and weekly impact reporting for CTR, conversion, and forecast accuracy.

• Built CI/CD with Jenkins and GitHub Actions: pytest coverage >85%, Docker multi-stage builds to ECR, and Helm deployments to EKS.

• Managed infrastructure as code with Terraform for SageMaker, Feature Store, Kinesis, EMR, and Elastic search with state in S3 and Dynamo DB locking.

Tech Stack: Python, Scikit-learn, XGBoost, Light GBM, PyTorch, Hugging Face Transformers (BERT, Sentence-BERT), Tensor Flow, Pandas, NumPy, AWS (S3, EMR, SageMaker — training/inference/Feature Store/Pipelines/Model Monitor, Lambda, Step Functions, Kinesis, Dynamo DB, RDS, Athena, Cloud Watch, ECR), Apache Spark (PySpark), Apache Kafka, Apache Airflow (MWAA), dbt, Snowflake, FastAPI, Docker, Kubernetes (EKS), ML flow, Weights & Biases, Optuna, SHAP, Redis, Elastic search, Terraform, GitHub Actions, Jenkins

Client: Hucon Technologies Location: Hyderabad, India Role: Data Engineer Duration: July 2017 – February 2020 Project Overview:

Served as a Data Engineer within a major Indian telecom operator's Analytics division, contributing to a subscriber analytics and churn prediction platform. Built big data pipelines processing 500M+ daily Call Detail Records (CDRs), a Hadoop/Spark historical data lake, and a real-time network monitoring pipeline utilizing Kafka and Spark Streaming. Executed the migration of legacy on premise infrastructure to AWS EMR, building initial ML feature engineering pipelines and optimized ETL workflows. Key Contributions & Responsibilities:

• Designed and developed scalable PySpark ETL pipelines on a 30-node Hadoop/YARN cluster to process over 500M+ Call Detail Records (CDRs) daily, parsing Ericsson Media Gateway data, transforming subscriber records, and loading curated datasets into Apache Hive ORC partitioned tables for downstream analytics.

• Built real-time telecom network monitoring solutions using Apache Kafka, Spark Structured Streaming, and PySpark, ingesting OSS/BSS performance counters, calculating rolling network KPIs, and generating near real-time degradation alerts for Network Operations Center (NOC) teams.

• Engineered advanced customer analytics and churn prediction features using PySpark, Window Functions, User Defined Functions (UDFs), rolling aggregations, statistical modeling, usage trend analysis, social network analytics, and customer experience metrics to support machine learning models.

• Managed and optimized enterprise Apache Hive data warehouses on HDFS, improving query performance through partition pruning, Apache Tez, Hive LLAP, ORC optimization, and workload tuning, significantly reducing large-scale analytical query execution times.

• Designed high-performance operational data stores using Apache HBase, implementing optimized row-key strategies for low-latency storage and sub-second retrieval of historical telecom KPI data supporting real-time operational dashboards.

• Developed enterprise Informatics Power Center ETL workflows to integrate billing, CRM, and subscriber data, consolidating customer master records and processing over 30 million daily prepaid recharge transactions into Hive staging environments.

• Supported enterprise cloud migration initiatives by migrating on-premises Hadoop workloads to AWS EMR, utilizing AWS Data Sync, Amazon S3, Spark on EMR, and S3A connectors to modernize data processing while significantly reducing infrastructure costs.

• Modernized legacy data warehouse architectures by migrating Apache Hive workloads to Amazon Athena, implementing the AWS Glue Data Catalog, optimizing Hive QL for Presto SQL, and replacing Apache Oozie workflows with Apache Airflow orchestration.

• Designed and implemented a scalable Amazon S3 Data Lake using Bronze, Silver, and Gold data layers with Parquet storage formats, integrating AWS Glue, Amazon Athena, and Amazon Redshift Spectrum for enterprise analytics and self- service reporting.

• Designed and optimized enterprise Amazon Redshift data warehouses using Kimball Star Schema, implementing efficient Distribution Keys (DISTKEY), Sort Keys (SORTKEY), fact tables, dimension tables, and query optimization techniques to accelerate analytical workloads.

• Developed real-time customer activity monitoring pipelines using Amazon MSK (Kafka), Amazon Kinesis Data Firehose, PySpark Structured Streaming, and Amazon Dynamo DB, enabling proactive customer engagement through automated inactivity detection and CRM alerting.

• Optimized enterprise Oracle PL/SQL batch processing using DBMS_PARALLEL_EXECUTE, parallel execution techniques, query optimization, and workload tuning to significantly improve large-scale data processing performance.

• Tuned Amazon Redshift performance through advanced SORTKEY, DISTKEY, workload management (WLM), compression encoding, and query optimization strategies, dramatically reducing analytical query execution times.

• Implemented enterprise data quality frameworks using Great Expectations, integrating automated validation rules, data profiling, schema validation, completeness checks, null validations, and business rule enforcement into Apache Airflow ETL pipelines.

• Developed automated data reconciliation, freshness monitoring, and SLA validation processes using SQL, Apache Airflow, and Amazon Cloud Watch, ensuring data accuracy, operational reliability, and proactive alerting for enterprise data platforms.

• Collaborated with Data Engineers, Data Scientists, DevOps engineers, Network Operations teams, and business stakeholders to deliver scalable, high-performance big data solutions using Hadoop, Spark, AWS, Kafka, and modern cloud- native data engineering best practices.

Tech Stack: Python, PySpark, Apache Spark (Hadoop — on premise and AWS EMR), Apache Hive, Apache HBase, Apache Kafka, SQL

(Oracle, MySQL, Hive QL, Spark SQL), Informatics Power Center, AWS (S3, EMR, Glue, Redshift, Lambda, Athena — growing cloud adoption), Apache Airflow, dbt (initial adoption), Tableau, Power BI, Git, Jenkins, Linux (Bash), Oracle Database (PL/SQL), Scikit-learn

(basic ML — feature engineering, model training), Elastic search Client: AstraZeneca Location: Hyderabad, India

Role: Software Engineer Duration: June 2015 – June 2017 Project Overview:

Served as a Software Engineer within an Indian private bank's Analytics and MIS division, supporting business intelligence, regulatory reporting, and data warehouse operations. Developed data engineering workflows utilizing Oracle SQL for data extraction, Python scripting for process automation, Informatics Power Center for ETL pipeline execution, and SSRS for management dashboards, while establishing automated data quality validation controls. Key Contributions & Responsibilities:

• Developed and optimized Oracle SQL queries against enterprise data marts to extract operational data for accounting, lending, and core banking systems, supporting business reporting, financial analysis, and regulatory reporting requirements.

• Utilized advanced Oracle SQL features including CASE, DECODE, ROW_NUMBER, LAG, analytical functions, and complex joins to transform and prepare data for monthly financial, operational, and compliance reporting.

• Designed, enhanced, and maintained operational reports using SQL Server Reporting Services (SSRS), developing interactive drill-through reports and dashboards to monitor product sales, ATM availability, and business performance metrics.

• Generated ad hoc SQL reports and data extracts to support statutory reporting, including Priority Sector Lending (PSL) analysis, regulatory submissions, and business intelligence requests.

• Assisted in maintaining enterprise Informatics Power Center ETL workflows, supporting incremental data loading, high- water-mark processing, data transformation, and core banking data integration into enterprise data marts.

• Executed, monitored, and troubleshot scheduled SQL Server Integration Services (SSIS) batch jobs, reviewing execution logs, resolving data validation issues, and supporting automated processing of CIBIL credit bureau data files.

• Maintained and validated Oracle PL/SQL stored procedures, packages, functions, and exception handling routines to support monthly loan processing, financial reporting, and operational data management.

• Developed automation scripts using Python, Pandas, and Open PyXL to transform raw text files into structured Microsoft Excel reports, eliminating manual data preparation and improving reporting efficiency.

• Performed data validation and quality checks within a Kimball Star Schema data warehouse by verifying dimension tables, fact tables, surrogate keys, and historical customer and branch data consistency.

• Utilized Git for source code management, version control, and change tracking of SQL scripts, SSRS reports, ETL components, and database objects while following established development standards.

• Participated in Agile Scrum ceremonies, including daily stand-ups, sprint planning, sprint reviews, and backlog discussions, collaborating with senior analysts, business users, and development teams to gather reporting requirements and deliver business intelligence solutions.

• Collaborated with database administrators, ETL developers, business analysts, and reporting teams to support enterprise data warehouse operations, regulatory reporting, and continuous improvement of data management processes. Tech Stack: SQL (Oracle — advanced, MySQL, SQL Server — intermediate), Python (intermediate — Pandas, NumPy, SQLAlchemy, Matplotlib — learning and applying), Informatics Power Center (ETL — learning and applying), SQL Server Reporting Services (SSRS), SQL Server Integration Services (SSIS), Microsoft Excel (advanced — PivotTables, Power Query, VLOOKUP, VBA macros), Tableau

(basic), Git (basic), Linux (Bash — basic)



Contact this candidate