Shashikiran Javvaji

Director / Principal Engineer II

📍 Denver, Colorado
✉️ shashikiran.javvaji@gmail.com

18+ years designing, building, and scaling distributed systems across embedded platforms, cloud infrastructure, and backend services, now focused on applied and agentic AI.

Request Resume LinkedIn

18+years engineering

US Patentholder

Millionsof devices served

15+years mentoring

About Me

Director / Principal Engineer focused on applied and agentic AI, backed by 18+ years building distributed systems across embedded, cloud, and backend platforms. At Charter's Emerging Technology group, I build AI-native engineering workflows and agent platforms on AWS: sandboxed evaluation of AI coding agents, self-improving code-review agents, multi-agent document pipelines, and Model Context Protocol (MCP) integrations that connect agents to enterprise tools.

I bring production rigor to AI adoption through measurable evaluation, guardrails, least-privilege access, and reproducible infrastructure as code, alongside 0-to-1 proofs of concept in latency measurement and network-as-a-service APIs. Previously a Staff Software Engineer at EchoStar/Dish, I delivered large-scale backend services, telemetry pipelines, and patented product features for millions of users.

Agentic AI

Hands-on work building, evaluating, and putting AI agents to work across the software development lifecycle.

AI Coding-Agent Evaluation Platform

Designed and built an isolated sandbox for evaluating AI coding agents on AWS. It covers infrastructure as code in AWS CDK, repeatable multi-instance deployments, a reproducible Nix toolchain, and controlled network egress, so agents reach only what they're allowed to.

Self-Tuning Code-Review Agent

Built an automated loop that rewrites an AI code reviewer's instructions, scores each revision against a graded test suite across multiple models, and keeps a change only when it's statistically better, or shorter with no loss in quality.

Multi-Agent Document Intelligence

Orchestrated a multi-agent workflow that inventories large enterprise document sets, converts mixed formats, and extracts structured facts tagged by source for human review.

Agents Connected to Enterprise Tools

Connected AI agents to a corporate wiki and issue tracker through approved Model Context Protocol (MCP) servers, scoped and rate-limited to fit company access policy.

Experience

Director / Principal Engineer II

Charter Communications (Emerging Technology)

Denver, CO

Jun 2023 – Present

Agentic AI platforms: Coding-agent evaluation, self-tuning review agents, multi-agent document pipelines, and MCP integrations (see Agentic AI above).AWS CDKBedrockMCP
AI-Augmented Development: Brought AI-native tools into design and implementation to speed POC delivery and cut boilerplate and debugging time.KiroCursorClaude Code
Latency Measurement: Led end-to-end design and deployment of a network-wide latency measurement platform, from field-trial firmware to cloud analytics.AWS IoT CoreMSK
Network-as-a-Service APIs: Built APIs that use real-time network and device data to help industries reduce fraud, authenticate users, and verify transactions.AWSAPIs

Staff Software Engineer

EchoStar Satellite LLC / Dish Technologies

Denver, CO

Jun 2008 – Jun 2023

Metadata Delivery: Built delivery pipelines for Dish, Sling, and web clients.AWS EKSArgoELK
Cloud Search Data: Built voice-search-ready search indices with automated data cleanup.ElasticsearchKubernetesDocker
Cloud Update: Replaced satellite targeting with over-the-air software updates for set-top boxes, cutting time and cost.Spring BootJavaPython
Real-Time Reporting: Live pipeline and dashboards for millions of set-top boxes to catch bugs before software reached the field.GrafanaJavaSpring BootPython
Android TV: Led the AirTV onboarding and remote-finder app with five partners, including Google TV.AndroidJavaPython
IPVOD UI: Rich animated UI on low-power set-top boxes for Bell Canada, avoiding hardware replacements.C
Screen Profiler & UI Skin Changer: Let customers switch UI themes and palettes without new software releases.CEmbedded
Direct Call Sign Entry: Patented feature letting viewers pick channels by name instead of number (US Patent 8,225,356).Patent
Multilingual Text Software: Patented graphics-compaction methods and multilingual UI for Dish, Bell, and TELUS.PatentLinux

Skills & Expertise

Languages

C
Java
Python
Shell Scripting

AI & Agentic Development

AWS Bedrock & AgentCore
MCP
Multi-agent orchestration
LLM evaluation
Kiro, Cursor, Claude Code
GitHub Copilot

Cloud Technology

AWS Bedrock, AgentCore
EC2, Lambda, IAM
MSK, Kinesis, Firehose, S3
Kubernetes
AWS CDK (infrastructure as code)

Web & Databases

JSON, HTML
MySQL
MongoDB

Data & Observability

ELK (Elasticsearch, Logstash, Kibana)
Grafana, OpenSearch
Splunk, Tableau

IDEs & Simulators

Eclipse, IntelliJ
Android Studio
Visual Studio, Atom

Operating Systems

Linux, macOS, Windows
Android, iOS
OpenWRT

Communication Protocols

TCP/IP, UDP, HTTP
MQTT, WebSocket
TR-069, DHCP, DNS

Packages & Tools

Git, ClearCase, Atlassian
Argo, Docker
Plume, OpenSync, Agora
NumPy, Pyshark, NetScout
Photoshop, Flash

Project Management

Agile
Scaled Agile (SAFe)
Waterfall

Engineering Leadership

Mentored interns and junior engineers for over 15 years at Charter Communications and Dish Network, helping them grow technically and professionally across embedded systems, firmware, and cloud infrastructure.
Coach engineers on working effectively with AI agents: writing specs and prompts, reviewing agent-generated code, and evaluating results against tests rather than intuition.
At Charter, provided hands-on mentorship for projects involving real-time network telemetry, latency measurement, and integration with analytics tools.
At Dish Network, guided junior engineers developing set-top box features, image processing enhancements, and multilingual UI implementations, and helped new engineers understand satellite communication system architecture and test automation practices.
Led by example in cross-functional collaboration, mentoring engineers across firmware, DevOps, and cloud services and helping them work effectively with platform, QA, and product teams to deliver reliable software.
Provided code reviews, architectural guidance, and documentation best practices across multiple platforms, fostering technical rigor and autonomy in mentees.
Built a culture of technical ownership by encouraging problem-solving, constructive feedback, and continuous improvement in engineering quality.

Education & Certifications

MS, Electrical Engineering

The University of Texas at El Paso (UTEP)

May 2008 · GPA 3.89 / 4.0

Machine Learning Certificate

Cornell University

Credential ID: kLKseZRMBc

Technology Leadership Certificate

Cornell University

Credential ID: EfQpCRdVwE

References available on request.

Let's Connect

Open to discussing new opportunities, technical challenges, or collaborations.

Email me LinkedIn