Saketh Reddy
I’m an AI backend engineer with 4+ years of experience building production cloud-native AI platforms, distributed Python microservices, and LLM-backed systems across AWS and Azure.
About
I’m an AI backend engineer with 4+ years of experience building production cloud-native AI platforms, distributed Python microservices, and LLM-backed systems across AWS and Azure. I’ve worked on multi-agent architectures, retrieval-augmented generation (RAG), vector search, and automated LLM evaluation to deliver reliable, governed AI features for real business workflows like underwriting, compliance automation, and advisor support.
I enjoy turning complex AI requirements into robust services—designing scalable APIs, observability, and CI/CD, and adding human-in-the-loop approvals and guardrails where it matters most. From predictive maintenance and anomaly detection to enterprise governance and knowledge-connected agent platforms, I focus on building dependable systems that perform well, are measurable, and integrate cleanly with existing data and enterprise tools.
- 4 years' experience
- Joined September 2026
Industry experience
- Financial Services
- Manufacturing
- Software & Internet
Work experience
Architected an enterprise AI platform for loan underwriting and compliance automation using multi-agent workflows with FastAPI, LangGraph, and Amazon Bedrock, reducing manual case processing by 50%. Implemented hybrid retrieval combining BM25 and OpenSearch vector search with reranking to improve grounded response accuracy from 74% to 92%. Built multi-agent orchestration for document analysis, compliance validation, and risk assessment with human-in-the-loop approvals, and established AI governance guidelines. Developed asynchronous Python microservices using Redis, Docker, and Kubernetes, with production observability via OpenTelemetry, Prometheus, and Grafana. Implemented automated LLM evaluation and prompt regression tests in LangSmith with pytest and GitHub Actions CI, including guardrails and PII/PHI masking for regulatory compliance.
Built a financial advisor copilot using FastAPI services that consolidated Salesforce, SQL Server, and DB2 into a single client profile API, reducing pre-meeting prep by 35%. Developed insurance document intelligence using Azure OpenAI and Cognitive Services to extract and validate fields from applications, cutting review time from two days to same day. Implemented semantic customer search over advisor records using LangChain and Azure AI Search embeddings, replacing exact-match lookups to improve relevance on first query. Delivered financial profile summarization and REST APIs on Azure SQL and Cosmos DB with connection pooling, reducing average API latency by 30% during peak advisor activity.
Built a manufacturing intelligence platform for predictive maintenance by creating Python feature pipelines and training Random Forest and Gradient Boosting models on Amazon SageMaker to reduce unplanned downtime by 27%. Designed telemetry-processing microservices with FastAPI and Apache Spark to ingest millions of sensor readings, adding Isolation Forest anomaly detection to flag abnormal equipment behavior early. Developed backend microservices exposing equipment-health scores and supply-chain optimization signals to operations dashboards, improving spare-parts planning and maintenance prioritization.
Education
Hire Saketh Reddy today
Post a job and invite Saketh, or contact them directly. No hidden commission.