I'm a Senior AI Engineer at Twilio, leading the development of Twilio Agent Connect - a model-agnostic orchestration layer that bridges AI agents with real-world communication channels. With a strong background in building scalable, production-grade systems, I'm passionate about creating innovative solutions that make a real impact.
- 🔭 Currently leading Twilio Agent Connect - featured at SIGNAL 2026 as a critical product launch
- 🌟 Top contributor to open source SDKs enabling multi-channel AI agents (Voice, SMS, WhatsApp, RCS)
- 💡 Previously built AI Assistants platform with RAG pipelines and Twilio Engage customer engagement features
- 🎯 Focused on delivering production-grade AI infrastructure and developer tools
🏢 Senior AI Engineer @ Twilio - Conversation AI Team
Current - Leading Twilio Agent Connect
Twilio Agent Connect - Model-agnostic orchestration layer connecting AI agents to multi-channel communications
- 🎯 Tech Lead & Architect: Designed and implemented the complete multi-channel, model-agnostic architecture enabling AI agents across Voice, SMS, WhatsApp, and RCS
- 🔧 Open Source Leadership: Lead contributor and maintainer of:
- Core TAC SDKs: Python and TypeScript - Building intelligent, context-aware AI agents across languages
- AWS Integration Package - Top contributor, simplifying AWS Bedrock, Nova, Claude, and Llama integration with Twilio
- 🏗️ Production Infrastructure: Architected streaming voice capabilities with barge-in detection, timeout handling, and sub-50ms latency WebSocket optimization
- 🤖 AI Tools & Features: Designed and implemented Memory Tools, Knowledge Base integration, and seamless AI-to-human handoff workflows
- 📊 AWS Partnership: Led the development of AWS connectors that dramatically simplify deploying multi-channel AI agents with Amazon Bedrock Agent, AgentCore, and foundation models
- 🚀 Critical Launch: Featured product at Twilio SIGNAL 2026, enabling organizations to bridge "great demo bots" to production communications
🏢 Senior Software Engineer @ Twilio - Emerging Tech Innovation (ETI)
2023 – 2024
AI Assistants - Opinionated framework for building and hosting customer-facing AI agents
- 🤖 RAG & Knowledge Management: Architected production-grade Knowledge Crawler indexing millions of URLs, websites, and files with automatic chunking and embeddings for real-time, context-aware responses
- 🏗️ Platform Architecture: Designed multi-tenant Control Plane and plugin-based orchestration layer enabling Assistants to execute tool calls and integrate with external APIs
- 💾 Customer Memory Integration: Built conversation history tracking with AI Perception and Personalization Engines, unified with Segment for cross-channel customer understanding
- ⚡ Infrastructure & Reliability: Managed Kubernetes infrastructure and CI/CD (Buildkite) with 99.9% availability via Datadog
- 🔒 Security & Guardrails: Implemented prompt injection protection, content moderation, and type-safe validation for enterprise-grade safety
🏢 Software Engineer @ Segment (Twilio)
Previous Role
Contributed to Twilio Engage - a leading customer engagement platform
- 📱 Built Mobile Push Notifications features enabling personalized messaging campaigns across iOS and Android platforms
- 📊 Developed Analytics and data visualization tools to help businesses track campaign performance and user engagement metrics
- 🏗️ Architected Organization and Permission Systems to manage multi-team workspaces with role-based access control
- ⚡ Optimized performance for high-throughput data processing, handling millions of events per second
- 🔧 Implemented features for audience segmentation, journey orchestration, and real-time personalization
🏢 Software Engineer @ BrightEdge
Previous Role
Worked on Autopilot - an AI-powered SEO platform
- 🤖 Implemented intelligent automation features for AI-driven content recommendations and SEO optimization workflows
- 🔧 Built tools for content optimization enabling marketers to improve organic search performance and rankings
- 📊 Developed data processing pipelines to analyze website performance metrics and competitor insights
- 🏗️ Enhanced platform scalability and reliability to support enterprise-level SEO operations
- ⚡ Integrated machine learning models for predictive analytics and automated decision-making in digital marketing
- Fine-Tune & Deploy LLMs with QLoRA on Sagemaker + Streamlit - Udemy (Oct 2025)
- Skills: Large Language Models (LLM), AWS SageMaker, PyTorch, TensorFlow
- The Complete Agentic AI Engineering Course (2025) - Udemy (Oct 2025)
- Skills: Large Language Models (LLM), Agents, OpenAI API, Python
- LLMs with Google Cloud and Python - Udemy (Dec 2023)
- Skills: Large Language Models (LLM), Python, Google Cloud Platform (GCP), Pandas, Jupyter
- Next.js 14 & React - The Complete Guide - Udemy (Jan 2024)
- Skills: React.js, Next.js, CSS, JavaScript, HTML, MongoDB
- CSS - The Complete Guide 2024 - Udemy (Mar 2023)
- Skills: SASS, CSS, Cascading Style Sheets, Web Applications
- GraphQL with React: The Complete Developers Guide - Udemy (Jan 2022)
- Skills: GraphQL
- Complete Guide to Protocol Buffers 3 [Java, Golang, Python] - Udemy (Aug 2024)
- Skills: Protocol Buffers
- Pydantic V2: Essentials - Udemy (Jul 2024)
- Go Design Patterns - LinkedIn (Apr 2022)
- Skills: Go (Programming Language)
- Learning the Go Standard Library - LinkedIn (Apr 2022)
- Skills: Go (Programming Language)
- Learn DevOps: Infrastructure Automation With Terraform - Udemy (Apr 2023)
- Skills: Terraform, Amazon Web Services (AWS)
- Datadog: Performance monitoring tool (from Zero to Hero) - Udemy (May 2022)
- Skills: Datadog
- Git & GitHub - The Practical Guide - Udemy (Jun 2023)
- YAML Zero to Master - Udemy (Mar 2023)
- Skills: YAML, Terraform
-
- Springer, Cham - May 2017
- Research on detecting community structure of brain networks using rs-fMRI data to determine differences between autism spectrum disorders (ASDs) and normal controls. Proposes GAcut method using genetic algorithm to automatically detect community structure based on optimized modularity Q.
-
Uncovering Community Structure in Neuronal Functional Networks from Multi-neuronal Spike Trains
- Springer, Singapore - January 2016
- Proposes a neuronal functional network community structure detection method using random walk distance and spectral decomposition to automatically determine the number and structure of neuronal functional networks.
-
Partitioning the Firing Patterns of Spike Trains by Community Modularity
- CogSci 2015 - July 2015
- Novel approach to analyze groups of firing patterns of neuronal spike trains based on community structure partitioning and modularity function Q, enabling automatic identification of optimal number of groups in neuronal firing patterns.
- 🎯 Leading Twilio Agent Connect - Featured at SIGNAL 2026, bridging AI agents to production communications at scale
- 🔧 Top Open Source Contributor - #1 contributor to Agent Connect AWS package, lead maintainer of Python & TypeScript SDKs
- 🏗️ Multi-Channel AI Architecture - Designed model-agnostic orchestration supporting Voice, SMS, WhatsApp, and RCS
- ⚡ AWS Integration Expert - Simplified Bedrock Agent deployment with production-ready connectors and sub-50ms latency optimization
- 💻 Production-Grade Systems - Built RAG pipelines, distributed systems, and customer engagement platforms serving millions
- 🚀 Innovation Track Record - From AI Assistants to Twilio Engage to Agent Connect, consistently shipping critical products
