Rithika Patil

AI Researcher

Bengaluru, Karnataka, India2 yrs 10 mos experience

Highly StableAI Enabled

Key Highlights

Developed a standardized platform for multi-agent AI systems.
Architected a banking assistant that reduced response time by 40%.
Created a RAG pipeline improving retrieval accuracy by 60%.

Stackforce AI infers this person is a highly skilled AI Engineer specializing in infrastructure and security for multi-agent systems.

Contact

Skills

Core Skills

Amazon Web Services (aws)Agent OrchestrationRuntime SecurityEnterprise Knowledge GraphGovernanceAi Banking AssistantAws IntegrationLegacy ModernizationSemantic Retrieval

Other Skills

LangChainOpenTelemetryLangSmithDatadogFastAPIDynamoDBAgentBoilerPlateAtomGuardConstitutional AILangGraphGovernance Summarisation AgentAmazon LexAmazon ConnectFAISS embeddingsCustom chunking

About

At Mphasis, I build the infrastructure that makes AI agents work in production. I have worked on building ATOM ecosystem, an internal platform that standardizes how teams build, observe, and secure multi-agent AI systems. It includes a rapid agent development framework, provider-agnostic observability (OpenTelemetry, LangSmith, Datadog), and runtime security - prompt injection protection, inter-agent trust, OWASP-aligned guardrails. Now adopted across Applied AI teams as the standard for building agent systems. Beyond platform work. I've built production systems including a multi-agent banking assistant (AWS Bedrock + Lex + Connect) that reduced response latency by 40%, and a RAG pipeline that modernizes legacy COBOL codebases with 60% better retrieval accuracy. What I care about: agent orchestration architectures, LLM observability, runtime security for AI systems, and making GenAI actually work at enterprise scale. Looking to work on AI systems where infrastructure, reliability, and scale are first-class concerns. You can reach me at rithikapatil213@gmail.com or directly here on LinkedIn.

Experience

2 yrs 10 mos

Total Experience

2 yrs 10 mos

Average Tenure

2 yrs 10 mos

Current Experience

Mphasis

AI Software Engineer

Aug 2023 – Present · 2 yrs 10 mos · Bengaluru

1. ATOM Ecosystem — Internal Agentic AI Platform Led development of ATOM, a standardized platform for secure, observable multi-agent AI systems. Core components: AgentBoilerPlate (cookiecutter framework with ReAct, multi-LLM, versioned prompts & memory), Pulse (observability across Opik, OpenTelemetry, LangSmith, Datadog), and AtomGuard (runtime security with prompt injection protection, trust policies, audit logging & OWASP-aligned fail-closed execution). Now the standard framework for agent development.
2. Ontologic — Enterprise Ontology Governance Platform Built a governed multi-agent platform for Enterprise Knowledge Graph creation using a 10-agent LangGraph pipeline, constitutional AI, and lifecycle state machine. Introduced Constitution Guardian (20 compliance checks with 15 hard VETOs), LLM-agnostic router with tiered model routing, and Governance Summarisation Agent that turns feedback into reusable context. Delivered full workflow (Draft → Peer Review → Board Vote → EKG) with immutable audits, boosting quality and reuse.
3. NeoMVerse — AI Banking Assistant Architected production multi-agent banking assistant with Amazon Lex, Bedrock, FastAPI/Lambda & DynamoDB. Built specialized agents for auth, balances, loans & payments. Added intelligent MCP-based LLM routing and Amazon Connect multilingual voice support. Reduced average response time by 40% (16.8s → 10.3s).
4. NeoZeta — GenAI Legacy Modernization Created RAG pipeline for semantic retrieval and modernization of undocumented COBOL systems. Implemented custom chunking and FAISS embeddings optimized for COBOL syntax. LLM workflows extract logic and generate documentation. Delivered 60% better retrieval accuracy and 50% faster developer onboarding.

Amazon Web Services (AWS)LangChainOpenTelemetryLangSmithDatadogFastAPI+2