Mentoring & Teaching
9 consecutive years in Google Summer of Code (2017-2026). 40+ students mentored across cloud-native architecture, multi-agent AI, distributed systems, and open-source development. I believe the best way to deepen your own engineering is to teach it.
Google Summer of Code (2017-2026)
Mentor (2019-2026) for Google Summer of Code, guiding 10+ students on cloud-native and systems projects in Go. Started as a contributor in 2017, became a mentor in 2019, and have mentored continuously through 2026. Supervised code quality, consulted on product ideation, architecture, and decision making. Fostered open-source culture and helped students build production-grade cloud-native systems.
- Mentoring Genie — an Agentic AI Financial Assistant built on MCP with tool calling, memory systems, and multi-agent orchestration
- Guided architecture decisions for 15 role-specialized agents (supervisor, analyzer, forecaster, anomaly_detector, recommender, llm_auditor)
- Teaching production patterns: JWT + RBAC, AES-256-GCM encryption, OpenTelemetry tracing
- Mentored RustCloud — a unified generative AI provider in Rust abstracting LLM providers with a consistent API for generation, streaming, and embeddings
- Guided students on Rust idioms, trait-based abstraction design, and provider abstraction patterns
- Consulted on product ideation and architecture for multi-provider LLM integration
- Selected as GSoC contributor for Gocloud — built unified API abstractions for multi-cloud resource management
- Contributed Go packages for compute, storage, and DNS across AWS, GCP, and Azure
- Successfully completed both 2017 and 2018 programs
Google Code-in (2017-2019)
Mentored pre-university students (ages 13-17) through Google Code-in, a global coding competition that introduced young developers to open-source contribution. Guided students through bite-sized tasks in Go, documentation, testing, and cloud tooling.
GCI Mentor — SCoRe Lab
Designed and reviewed tasks for pre-university students across 3 consecutive years. Tasks spanned Go programming, API design, documentation, and testing — introducing young developers to professional open-source workflows.
Team & Engineering Mentoring
Beyond open-source programs, I've mentored engineers and interns in professional settings — setting architecture standards, running code review, and building engineering culture.
Engineering Mentor — Searce
Mentored 40+ developers across 4 internship cohorts. Set architecture standards across product lines. Ran structured code reviews, pair programming sessions, and architecture workshops for cloud-native and AI systems.
Team Lead — Globe Telecom Project
Led a team of 10 engineers building a 30K+ TPS Kubernetes transaction platform. Coached junior engineers on Go concurrency patterns, PCI compliance, and distributed systems design through daily pairing and architecture reviews.
Remote Team Lead — Picnic Social Platform
Led a remote-first international team of 4 engineers. Established async code review culture, architecture decision records, and 80%+ automated test coverage standards across the Go/gRPC/GraphQL backend.
Open Source Collaboration — AT&T, Ericsson, Microsoft
Contributed to airshipit and azure-service-operator in collaboration with engineers from AT&T, Ericsson, and Microsoft. Cross-organization code review and architectural alignment.
Teaching Through Writing
103+ published technical articles covering multi-agent AI, cloud architecture, security, governance, and production systems. Every post cites working code and is written to teach, not just inform.
ADK to MARA Migration Series
8-part deep-dive series documenting the migration of 18+ agents from Google ADK to Microsoft MARA. Covers orchestration patterns, provider abstraction, tool governance, state management, and deployment.
Security & Governance Deep-Dives
Long-form posts on defence-in-depth for agentic AI, HIPAA-as-Go-interfaces, RBI FREE-AI implementation, PAM analogs, and compliance-as-code. Each maps regulatory requirements to specific file paths in working codebases.
Production Systems Case Studies
Detailed write-ups on real production systems: BigQuery FinOps (57% cost reduction), Globe Telecom (30K TPS), Spanner migrations, and Kubernetes transaction platforms. Written as teaching material, not marketing.
Open Source as Teaching
Both Genie and Bodh are designed as reference implementations — not just working software, but codebases that teach multi-agent architecture patterns by example.
Genie — Multi-Agent Financial Assistant
Go-based reference implementation of Microsoft MARA with 15 role-specialized agents. Designed so GSoC students and contributors can learn production patterns: JWT + RBAC, AES-256-GCM encryption, OpenTelemetry, and multi-agent orchestration.
Bodh — Medical AI Platform
Original open-source medical AI platform on MARA inspired by Microsoft's MAI-DxO. Role-specialized agents for clinical workflow. Designed to teach FHIR R4, HL7 v2 integration, cost-aware diagnostic budgeting, and healthcare compliance patterns.
Mentoring Philosophy
- Teach architecture decisions, not just syntax — every code review is a conversation about trade-offs
- Write reference implementations that are readable first, performant second — code that teaches is code that lasts
- Document honestly — mark what's done, what's partial, and what's a genuine gap
- Open source is the highest-leverage form of teaching — one well-designed codebase reaches more engineers than any workshop
Related Writing
Read about the lessons, patterns, and trade-offs from 9 years of open-source mentorship and production engineering.