Hi, I’m Jaison Felix Menezes

Software Engineer, Co-op · Ancestry · Lehi, UT

Feb 2026 – Present · Python · OpenTelemetry · Model Context Protocol (MCP) · LangChain · Claude API · AWS · CloudWatch

• Developing AI-powered migration tools using LangGraph and Claude API.
• Building agentic Python AST parsing engines using Model Context Protocol (MCP) to migrate New Relic and Datadog APM instrumentation to OpenTelemetry with automated validation and rollback safety.
• Implemented real-time PII detection and redaction pipelines using AWS Lambda and CloudWatch Logs to scan and sanitize multi-source log streams for compliance.

Business Intelligence Engineer, Intern · Amazon · Bellevue, WA

May 2025 – Aug 2025 · SQL · Spark · Redshift · Kinesis · Lambda/Glue · Airflow

• Built a labor-optimization service that disaggregates shift schedules into hourly load forecasts — improved granularity and reduced overstaffing by 14% at a pilot site in 6 weeks.
• Implemented real-time event processing (Kinesis → Lambda/Glue → Redshift) with durable storage & query APIs; refreshed 7 KPIs in 8–12 minutes end-to-end vs. 2–3 hours baseline.
• Productionized A/B evaluation for a First-Mile pre-bagging pilot; ETL (Redshift SQL, Airflow) for touch count/mis-sorts/truck fill rate contributed to ~8% fewer touches.

Research Assistant · CSULB Research · Long Beach, CA (Part‑time)

Jan 2025 – Feb 2026 · Python/Flask · Java/Kotlin · Apache Spark (Scala) · Iceberg/Delta

• Shipped a WCAG 2.2 AA compliant, parent‑facing Flask app with ARIA attributes and automated accessibility checks.
• Refactored Java microservices (auth, payments) for distributed deployment; +JUnit/Mockito tests cut bug resolution time by 25%.
• Architected ingestion (Databricks Workflows + Spark) moving 20+ GB/day of Census/IRS/USPS data into Iceberg ACID tables; later migrated to Delta for compaction efficiency.
• Tuned a Kotlin Android app with live class schedules & charts (MPAndroidChart), adopted by 1,000+ users.

Software Engineer — Data & ML Infrastructure · ICICI Lombard · Mumbai, India

Jan 2023 – Aug 2024 · PySpark · Databricks · Azure · Cosmos DB · Redis · APIM

• Engineered batch + streaming services landing in ADLS Gen2/Delta across 5+ sources (~1.2M records/month); 99.9% successful runs, DLQ <0.1%.
• Real‑time telematics via Event Hubs (Kafka API) → Cosmos DB; sustained 1.2–1.5K msgs/s, p95 read ~5 ms, p99 write ~30 ms.
• Fraud‑scoring REST APIs (RF/XGBoost/LogReg) on Azure Functions behind APIM (rate limits, OAuth2/JWT); manual claim review time −40%.
• Distributed caching (Redis + Bloom filters) across 15+ microservices; DB load −60%, API p95 from 250 ms → 70 ms.
• Provisioned infra with Terraform; Jenkins caching cut CI/CD times by 40%.

Software Engineer · CloudFront Technologies (DataFronts) · Jersey City, NJ (Remote)

Aug 2020 – Jan 2023 · Java (Spring Boot) · Azure App Service · RabbitMQ · Azure DevOps

• Built event‑driven microservices integrating CRM (Salesforce/Dynamics) & ERP (SAP/NAV/F&O); zero‑downtime releases via staged slots.
• Personalization graph in Neo4j + Redis cache → sub‑100 ms recommendations; monitored with Azure Log Analytics.
• Nightly ERP→CRM sync on AKS (Docker, K8s, Helm): 2M+ rows/run, success 85%→99.5%, runtime −40%.
• Implemented API Gateway controls (APIM rate limiting + throttling) and secure auth (OAuth2, TLS); served 2,000+ active users.
• DRI for production data services; led on‑call rotations, incident triage, and SLA restoration.

FinReport AI

Python · RAG · Hugging Face · PostgreSQL · Docker

• NLP analytics on SEC filings using BERT/GPT with retrieval-augmented generation.
• Containerized API with dynamic model routing and prompt tuning.

View on GitHub

Yahoo Finance MCP Server

Python · MCP · yfinance · Claude API

• Built a custom Model Context Protocol (MCP) server that integrates Claude Desktop with live Yahoo Finance market data.
• Exposed production-ready tools including real-time stock price retrieval, historical CSV export, and multi-stock comparison.
• Designed structured tool interfaces over stdin/stdout MCP transport enabling Claude to programmatically query financial data.
• Packaged using modern Python tooling (uv, pyproject.toml), with modular server architecture for maintainability.

View on GitHub

GitHub MCP Server – Intelligent Repository Automation via Claude

Python · MCP · PyGithub · uv · REST API · Automation · Claude Desktop

• Built a custom Model Context Protocol (MCP) server that integrates Claude Desktop directly with the GitHub API for intelligent repository interaction and automation.
• Implemented secure GitHub authentication using Personal Access Tokens loaded via environment variables with python-dotenv.
• Developed tools to fetch user bios, repository file structures (recursive), and raw file contents dynamically from any GitHub repository.
• Enabled automated branch creation, pull request generation, and issue creation using the PyGithub SDK with structured error handling.
• Configured production-ready local execution using uv project environments to ensure dependency isolation and seamless integration with Claude Desktop.
• Designed the system to support scalable repository analysis, automation workflows, and AI-assisted codebase exploration.

View on GitHub

MovieHub

React · Node.js · MongoDB · Redis · Docker

• Implemented an ML-based movie recommender using cosine similarity on genre, keywords, and metadata vectors.
• Integrated a Gemini-powered AI assistant for contextual movie insights, containerized backend services with Docker, and deployed a scalable architecture (Vercel frontend, Render backend, MongoDB Atlas) with production-ready routing, security, and performance tuning.

GitHub Repo Live Demo

Azure Real-Time Weather Analytics & Alerting System

Azure · Databricks · Functions · Event Hub · Microsoft Fabric · KQL · Power BI

• Built an end-to-end real-time weather data pipeline that continuously ingests, streams, and analyzes weather data from a public Weather API.
• Implemented scheduled ingestion using Azure Functions and streaming pipelines with Azure Databricks into Azure Event Hub.
• Processed streaming events in Microsoft Fabric Event Stream and stored them in KQL (Kusto) DB for real-time querying.
• Designed a live Power BI dashboard delivering continuously updated analytics and alerting insights.
• Secured secrets and access using Azure Key Vault and Managed Identity with role-based access control (RBAC).

View on GitHub

Languages

Data & Storage

Pipelines & Orchestration

Machine Learning

Cloud & Infra / Security

CI/CD & Quality

Llama Hackathon & Meta — Participant (2025)

Built WikiFace, a DeepWiki‑inspired platform using Llama‑3 long‑context + chatbot UI to extract and centralize model insights; fine‑tuned on AWS p5s to reduce hallucinations; praised by peers & judges.

Hackniche 2.0 — Judge (Google DSC, 2024)

Evaluated projects from 1,300+ participants (teams of four) on scalability, creativity, and technical execution.