About Rupali
English
Native or bilingual
Hindi
Native or bilingual
Experience
- Finjo (AI Recruitment Platform)Senior Software EngineerJanuary 2023 - Today (3 years and 5 months)• • Building a production RAG pipeline for intelligent candidate–job matching using OpenAI embeddings + pgvector (PostgreSQL vector extension) served via Postgraphile GraphQL — reduced irrelevant shortlisting by ~40%.• • Designed vector search architecture — evaluated pgvector vs Pinecone vs Weaviate; implemented hybrid semantic + keyword search to handle edge cases where pure embedding similarity fell short.• • Architected fully serverless backend on AWS Lambda + API Gateway + SES; optimised cold-start latency to under 800 ms through careful function decomposition and payload tuning.• • Automated end-to-end candidate communication workflows (outreach, scheduling, status updates), eliminating ~15 hours/week of manual recruiter effort per client.• • Implemented security controls: JWT auth, PostgreSQL row-level security, AWS Secrets Manager — passed first external security audit with zero critical findings.• • Owned infra observability via CloudWatch dashboards + custom alerting; reduced MTTD on production incidents from hours to under 10 minutes.• • Stack: Python◦ · PostgreSQL◦ · pgvector◦ · LangChain◦ · OpenAI API◦ · Postgraphile◦ · GraphQL◦ · AWS Lambda/SES/API Gateway◦ · Docker
- Stealth StartupBackend & AI Integration EngineerMarch 2026 - Today (3 months)• • Design and deploy multi-agent systems using LangChain + LangGraph — agents handle decision branching, tool use (web search, DB queries, API calls), and handoff logic between sub-agents.• • Conduct workflow audits for enterprise clients: ingest SOPs, meeting transcripts, and process docs through an LLM pipeline to surface bottlenecks — clients report 30–50% reduction in operational review time.• • Architected a three-layer agent memory system: short-term (in-context), episodic (vector store retrieval via pgvector), and structured (PostgreSQL) — enabling persistent cross-session context.• • Evaluated multiple LLM providers (GPT-4o, Claude 3.5, Mistral via Ollama) for optimal cost/latency/accuracy trade-offs per task criticality.• • Stack: Python◦ · LangChain◦ · LangGraph◦ · OpenAI API◦ · Anthropic API◦ · pgvector◦ · FastAPI◦ · Docker
- Nutanix Technologies (BEAM SaaS)Member of Technical StaffApril 2018 - April 2021 (3 years)Bengaluru, Karnataka, India• • Built and maintained microservices for cloud cost analysis across AWS, Azure, and GCP — services processed millions of cost-data events per day at peak load.• • Owned Chargeback feature end-to-end: data model, REST APIs, frontend collaboration — shipped to 50+ enterprise customers, enabling cost breakdown by cost center / team / project.• • Reduced RDS query costs by ~25% via slow-query analysis, targeted indexing, N+1 ORM refactoring, and moving heavy aggregations to off-peak scheduled jobs.• • Improved observability with structured logging pipelines and custom monitoring — on-call escalation frequency dropped measurably quarter over quarter.• • Stack: Java◦ · Python◦ · PostgreSQL◦ · REST APIs◦ · AWS (RDS, EC2, S3)◦ · Microservices
Recommendations
Be the first to recommend Rupali
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- eMasters in CybersecurityIIT Kanpur2026eMasters in Cybersecurity
- B.Tech in Information TechnologyVellore Institute of Technology2018B.Tech in Information Technology