Ajith Surakanti

I'm a Software Engineer passionate about building production-grade AI systems and scalable backend platforms.

Over the past few years, I've worked on enterprise applications serving thousands of users, designing event-driven architectures, optimizing backend services, and building reliable, high-performance systems with Java, Spring Boot, PostgreSQL, and cloud technologies.

My current focus is Generative AI and Agentic AI. I enjoy building intelligent systems powered by LLMs, Retrieval-Augmented Generation (RAG), multi-agent workflows, knowledge graphs, and AI orchestration frameworks such as LangChain and LangGraph. I believe the future of software lies in systems where humans define goals and AI agents execute, collaborate, and continuously improve.

I'm particularly interested in

• Enterprise AI platforms

• AI agents and autonomous workflows

• RAG and semantic search

• LLM evaluation and observability

• Distributed systems and backend architecture

• Developer productivity and AI-assisted engineering

I enjoy learning by building—from production enterprise software to experimental AI projects—and I'm always exploring new approaches to make AI systems more reliable, scalable, and useful.

I'm open to connecting with engineers, researchers, and builders working on AI infrastructure, intelligent software, and large-scale distributed systems.

Software Engineer

UBS | Pune, India

Aug 2025 – Present

Backend Engineering

• Developed enterprise-scale backend services using Java 17, Spring Boot, PostgreSQL, Azure, and event-driven architecture for applications used by thousands of business users.

• Built and owned the Document Management module, implementing secure upload, retrieval, deletion, pagination, metadata management, and Azure Blob Storage integration.

• Designed and implemented 17+ event-driven and scheduled email notification workflows, automating critical business processes.

• Led the synchronization of Flowable attachments and metadata to Azure Blob Storage, including a one-time migration for legacy data.

• Optimized critical backend APIs, reducing response times by up to 70% through SQL optimization, query tuning, and performance improvements.

• Designed and implemented a configurable Audit Logging Framework to capture user actions and business events across multiple enterprise modules.

• Built features for approvals, milestones, conditions, case teams, reporting, and document management while ensuring scalability through asynchronous event processing.

• Contributed to enterprise workflow and approval data migration from legacy systems, implementing complex transformation logic and ensuring data integrity.

AI Engineering

• Developing an Agentic Case Management System that enables users to interact with enterprise cases using natural language and AI-assisted workflows.

• Built enterprise Retrieval-Augmented Generation (RAG) pipelines to provide contextual, source-grounded answers using case data, project documentation, requirements, and business knowledge.

• Evaluated and integrated Small Language Models (SLMs) and Large Language Models (LLMs), selecting the optimal model based on latency, reasoning capability, and operational cost.

• Designed multi-agent workflows using LangGraph/LangChain, enabling specialized agents for planning, retrieval, reasoning, document analysis, and response generation.

• Implemented hybrid retrieval by combining semantic search, metadata filtering, keyword search, and reranking to improve retrieval precision and response quality.

• Developed knowledge graph–based retrieval to model relationships between cases, approvals, requirements, documents, and other enterprise entities, enabling richer contextual reasoning.

• Integrated multiple LLM providers, embedding models, conversation memory, prompt management, AI tracing, and evaluation frameworks to build reliable enterprise AI applications.

• Built AI-powered enterprise search and copilot capabilities to improve knowledge discovery, document understanding, and productivity within case management workflows.

Technologies: Java 17, Spring Boot, PostgreSQL, React, Azure, Azure Blob Storage, Flowable CMMN, REST APIs, Event-Driven Architecture, LangChain, LangGraph, LLMs, SLMs, RAG, Hybrid Search, Knowledge Graphs, Vector Databases, Prompt Engineering, AI Evaluation, Azure DevOps, Git

B.Tech – Computer Science and Engineering

Visvesvaraya National Institute of Technology, Nagpur

Dec 2021 – May 2025