Jaya

🥑

[rag]knowledge base········OK

[vec]bge-base · ONNX········OK

[llm]gemini-2.5-flash········LIVE

[sys]avocado online········✦ ready

🥑

[rag]knowledge base········OK

[vec]bge-base · ONNX········OK

[llm]gemini-2.5-flash········LIVE

[sys]avocado online········✦ ready

Jaya

Career · Timeline

Experience

3+ years building production AI & distributed systems.

Currently active6 roles5 companies

ML Software Engineer
NYU CAS - Center for Genomics and Systems Biology
Jun 2025 – Present
New York, NY
LangGraphMachine LearningPyTorchPythonAWSFastAPIReactTypeScriptPostgreSQL
- Building GeneCart, an AI-assisted genomics discovery platform, owning the agent/LLM layer with LangGraph and PyTorch to make genomic data exploration actionable for researchers.
- Designing the full-stack delivery path, FastAPI services, a React/TypeScript frontend, and PostgreSQL on AWS and Kubernetes.
Software Engineer - Research Infrastructure
NYU IT – High-Speed Research Network
Sep 2024 – May 2025
New York, NY
PythonLangGraphLlama 3.1FastAPINode.jsAWS ECSKubernetesRedisTypeScript
- Architected a production Multi-Agent Research Discovery Engine (LangGraph + Llama 3.1 70B) that semantically maps global researcher collaboration networks by indexing millions of papers from Elsevier's (Science Direct/Scopus), enabling cross-institutional discovery at a scale previously requiring manual curation.
- Improved scientific query accuracy by 40% by engineering a Multi-Agent Search Router with hybrid retrieval, combining BGE-M3 dense embeddings with Reciprocal Rank Fusion (RRF) to resolve ambiguous cross-domain nomenclature lookups.
- Sustained 99.9% uptime at 3,000+ RPS by building a horizontally scaled async pipeline (Node.js + FastAPI on AWS ECS) with fault-isolated workers and circuit-breaker patterns.
- Reduced P99 retrieval latency by 78% (450ms to <100ms) by designing a Write-Through Redis caching strategy for multi-hop Chain-of-Retrieval queries across chained LangGraph agent calls.
- Cut deployment lead times by 65% by owning the full CI/CD lifecycle, containerizing stateful LLM workloads on Kubernetes with zero-downtime rolling deploys.
Software Engineer Intern
NYU IT – High-Speed Research Network
May 2024 – Aug 2024
New York, NY
FastAPIAWS ECS (Fargate)AsyncIORedisDockerKubernetes
- Built a FastAPI microservice on AWS ECS (Fargate) using AsyncIO to parallelize data fetching from 5+ external APIs, slashing P99 latency from 450ms to <100ms.
- Deployed a Redis Write-Through caching layer with custom TTL eviction policies, handling >3,000 RPS with sub-millisecond retrieval times.
- Modernized CI/CD pipelines using Docker and Kubernetes, enabling zero-downtime blue-green deployments.
TA – Data Structures & Algorithms / Machine Learning for Bioinformatics
NYU (Teaching Assistant)
Sep 2024 – May 2025
New York, NY
PythonJupyterML pipelines
- Mentored 50+ graduate students on Distributed System design patterns and algorithmic efficiency (Big O optimization); conducted code reviews for scalable ML pipeline implementations.
- Collaborated with faculty to design assignments on real-world ML workflows including data preprocessing pipelines and model evaluation metrics.
- Wrote Python scripts to automate grading of Jupyter notebooks, reducing feedback turnaround time by 40%.
Software Engineer
Wipro Limited (Client: Shell PLC)
Jul 2021 – Aug 2023
Bengaluru, India
JavaApache KafkaSalesforceApexLWCPostgreSQLAWS S3
- Delivered zero data-loss maritime telemetry for Shell PLC by architecting a fault-tolerant Apache Kafka pipeline processing 115GB/day of extraction and shipment data across 200+ offshore crude oil stations in Brazil.
- Reduced support ticket volume by 35% across 500+ vendors by launching a Self-Service Portal with automated credential lifecycle management and real-time KPI dashboards.
- Increased system fault tolerance by 39% by decomposing a legacy monolithic Salesforce integration into event-driven Apex Async Micro-functions, eliminating cascading failures.
- Reduced production incidents by 25% by engineering a defensive ETL validation layer that quarantines malformed payloads to an S3 Dead Letter Queue.
- Engineered automated workflows for Shell's Self-Help Portal, integrating cross-platform APIs to reduce ITIL service requests by 45%.
Full Stack Developer
Vellore Institute of Technology
Jan 2020 – Jun 2021
Vellore, Tamil Nadu, India
PHPMySQLJavaScriptSEO
- Engineered a PHP/MySQL backend that scaled to handle 17K+ monthly active users and 20K+ registered accounts.
- Achieved a #2 Google Search ranking via advanced programmatic SEO and optimized frontend rendering for sub-second load times on mobile networks.
- Designed normalized database schemas to ensure data integrity for high-volume grade calculations.

🥑

[rag]knowledge base········OK

[vec]bge-base · ONNX········OK

[llm]gemini-2.5-flash········LIVE

[sys]avocado online········✦ ready

Career · Timeline

Experience

3+ years building production AI & distributed systems.

Currently active6 roles5 companies

ML Software Engineer
NYU CAS - Center for Genomics and Systems Biology
Jun 2025 – Present
New York, NY
LangGraphMachine LearningPyTorchPythonAWSFastAPIReactTypeScriptPostgreSQL
- Building GeneCart, an AI-assisted genomics discovery platform, owning the agent/LLM layer with LangGraph and PyTorch to make genomic data exploration actionable for researchers.
- Designing the full-stack delivery path, FastAPI services, a React/TypeScript frontend, and PostgreSQL on AWS and Kubernetes.
Software Engineer - Research Infrastructure
NYU IT – High-Speed Research Network
Sep 2024 – May 2025
New York, NY
PythonLangGraphLlama 3.1FastAPINode.jsAWS ECSKubernetesRedisTypeScript
- Architected a production Multi-Agent Research Discovery Engine (LangGraph + Llama 3.1 70B) that semantically maps global researcher collaboration networks by indexing millions of papers from Elsevier's (Science Direct/Scopus), enabling cross-institutional discovery at a scale previously requiring manual curation.
- Improved scientific query accuracy by 40% by engineering a Multi-Agent Search Router with hybrid retrieval, combining BGE-M3 dense embeddings with Reciprocal Rank Fusion (RRF) to resolve ambiguous cross-domain nomenclature lookups.
- Sustained 99.9% uptime at 3,000+ RPS by building a horizontally scaled async pipeline (Node.js + FastAPI on AWS ECS) with fault-isolated workers and circuit-breaker patterns.
- Reduced P99 retrieval latency by 78% (450ms to <100ms) by designing a Write-Through Redis caching strategy for multi-hop Chain-of-Retrieval queries across chained LangGraph agent calls.
- Cut deployment lead times by 65% by owning the full CI/CD lifecycle, containerizing stateful LLM workloads on Kubernetes with zero-downtime rolling deploys.
Software Engineer Intern
NYU IT – High-Speed Research Network
May 2024 – Aug 2024
New York, NY
FastAPIAWS ECS (Fargate)AsyncIORedisDockerKubernetes
- Built a FastAPI microservice on AWS ECS (Fargate) using AsyncIO to parallelize data fetching from 5+ external APIs, slashing P99 latency from 450ms to <100ms.
- Deployed a Redis Write-Through caching layer with custom TTL eviction policies, handling >3,000 RPS with sub-millisecond retrieval times.
- Modernized CI/CD pipelines using Docker and Kubernetes, enabling zero-downtime blue-green deployments.
TA – Data Structures & Algorithms / Machine Learning for Bioinformatics
NYU (Teaching Assistant)
Sep 2024 – May 2025
New York, NY
PythonJupyterML pipelines
- Mentored 50+ graduate students on Distributed System design patterns and algorithmic efficiency (Big O optimization); conducted code reviews for scalable ML pipeline implementations.
- Collaborated with faculty to design assignments on real-world ML workflows including data preprocessing pipelines and model evaluation metrics.
- Wrote Python scripts to automate grading of Jupyter notebooks, reducing feedback turnaround time by 40%.
Software Engineer
Wipro Limited (Client: Shell PLC)
Jul 2021 – Aug 2023
Bengaluru, India
JavaApache KafkaSalesforceApexLWCPostgreSQLAWS S3
- Delivered zero data-loss maritime telemetry for Shell PLC by architecting a fault-tolerant Apache Kafka pipeline processing 115GB/day of extraction and shipment data across 200+ offshore crude oil stations in Brazil.
- Reduced support ticket volume by 35% across 500+ vendors by launching a Self-Service Portal with automated credential lifecycle management and real-time KPI dashboards.
- Increased system fault tolerance by 39% by decomposing a legacy monolithic Salesforce integration into event-driven Apex Async Micro-functions, eliminating cascading failures.
- Reduced production incidents by 25% by engineering a defensive ETL validation layer that quarantines malformed payloads to an S3 Dead Letter Queue.
- Engineered automated workflows for Shell's Self-Help Portal, integrating cross-platform APIs to reduce ITIL service requests by 45%.
Full Stack Developer
Vellore Institute of Technology
Jan 2020 – Jun 2021
Vellore, Tamil Nadu, India
PHPMySQLJavaScriptSEO
- Engineered a PHP/MySQL backend that scaled to handle 17K+ monthly active users and 20K+ registered accounts.
- Achieved a #2 Google Search ranking via advanced programmatic SEO and optimized frontend rendering for sub-second load times on mobile networks.
- Designed normalized database schemas to ensure data integrity for high-volume grade calculations.

ML Software Engineer

Software Engineer - Research Infrastructure

Software Engineer Intern

TA – Data Structures & Algorithms / Machine Learning for Bioinformatics

Software Engineer

Full Stack Developer

ML Software Engineer

Software Engineer - Research Infrastructure

Software Engineer Intern

TA – Data Structures & Algorithms / Machine Learning for Bioinformatics

Software Engineer

Full Stack Developer