Help & Supportcareers@gravityer.com
Software Engineer with 2+ years in AI-powered Applications & Cloud Systems
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Applied AI Engineer with 2+ years building AI-powered applications and operational workflows across enterprise and personal projects. Designed multi-agent systems, RAG pipelines, and evaluation workflows with hands-on experience in FastAPI, LangChain, LangGraph, vector search, observability, and production readiness. Focused on building grounded, secure, and operationally reliable AI systems.
SASTRA Deemed University
Bachelor of Technology · Computer Science
August 1, 2019 – June 1, 2023
Optum
Software Engineer
September 1, 2023 – Present
Hyderābād, Telangana, India
LegalMind AI - Production-Ready RAG System for Legal Document Intelligence
June 1, 2023 – Present
Designed a LangGraph ReAct agent with 5 specialized tools (document Q&A, clause summarization, entity extraction, compliance review, cross-document comparison); SSE-streamed full Thought/Action/Observation traces to a React frontend for auditable reasoning. Built a 5-stage async ingestion pipeline: PyMuPDF/python-docx parsing → chunking → Sentence Transformers embeddings → per-user Qdrant collection → GCS storage; Cloud Tasks for background processing with retry and stale-job recovery. Evaluated RAG pipeline with RAGAS across faithfulness, context precision, and answer relevancy; iteratively tuned chunking strategy and retrieval parameters based on metric feedback. Implemented production-grade AI safety layer: prompt-injection resistance, sensitive-data detection, weak-evidence warnings, legal disclaimers, and document-scope enforcement.
View ProjectFinancial News AI - Full-Stack GenAI Financial Intelligence Platform
June 1, 2023 – Present
Orchestrated a 4-agent ADK pipeline (News Analyzer → Analyst → Insight → Comparison Agent) with Gemini API for sentiment classification (positive/neutral/negative), confidence scoring, and BUY/HOLD/SELL signal generation. Applied 3-hour Firestore caching strategy to reduce LLM inference calls; per-user Firebase Auth and Firestore history for session isolation. Deployed FastAPI on Cloud Run and React on Vercel; full-stack production architecture with agent-level observability.
View ProjectHealthcare Insurance Claim Approval Agent
June 1, 2023 – Present
Built a LangGraph ReAct agent with 3 clinical tools (record summarization, policy parsing, claim evaluation); outputs structured APPROVE/ROUTE decisions enriched with ICD-10 and CPT codes for full clinical explainability. Applied LLM-as-a-Judge evaluation methodology across 10 human reference cases; iteratively refined system and tool-level prompts to minimize hallucination and maximize decision accuracy.
Microsoft Certified: Azure Fundamentals (AZ-900)
Microsoft
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's projects demonstrate a diverse application of AI in different domains (legal, finance, healthcare, IT operations), indicating adaptability and a broad interest in problem-solving with AI. The combination of enterprise experience at Optum and personal projects shows initiative and a continuous learning mindset. The focus on building robust, secure, and observable systems aligns well with a culture that values quality and operational excellence.
Soft Skills & Operational Fit
The candidate's project descriptions and professional summary indicate a strong focus on building 'production-ready,' 'grounded,' 'secure,' and 'operationally reliable' AI systems, suggesting a good fit for roles requiring attention to detail, reliability, and practical application of AI. The emphasis on evaluation methodologies (RAGAS, LLM-as-a-Judge) also points to a data-driven and iterative approach to development.