Drag & drop each image into the matching slot — they save in this browser and appear everywhere on the portfolio immediately.
// stored locally · no server upload · persists across page reloads
Leading Cortex Code AI integration, governed Cortex Agents over Snowflake, and MCP server architecture for natural-language analytics across healthcare & retail clients.
Lead architect on VeriTracks — AWS + Snowflake real-time device platform. 300K devices · 83 TPS · −83% run cost. Authored ARB TACR; led 4 AI audit cycles to ~95% legacy parity.
Reusable ELT framework on Snowflake + Python. SCD1/2 merges, audit + rejection-handling, multithreaded DB2 OMS ingestion, OAuth + key-pair auth.
SQL Server / MySQL / Redshift → Snowflake with OpenFlow CDC, PHI/PII masking, Tableau modernization. +60% query perf. HIPAA aligned.
Hybrid AWS + Azure data platform; Databricks + PySpark; custom Kafka Connect to 15+ sources; Jenkins + Docker CI/CD; CloudWatch + Airflow observability.
Azure Databricks + Data Factory ETL platforms; batch and streaming PySpark workflows. Star Performer + Excellence in Data Engineering awards.
Built front and back-end of Person Verification project with Python + Django; API integrations; pytest unit testing.
Master of Science with focus on data science, machine learning, distributed systems, and cloud-native architecture.
// portrait · 2026
Architecting Snowflake-native + AWS-native systems. Cortex Code, Anthropic Claude, governed RAG, MCP servers, token budgeting.
Anthropic Claude, Snowflake Cortex Code, AWS — production AI workflows with governance and cost discipline.
Leading AI & data-platform modernization at enterprise scale across healthcare, retail, fintech, public-safety.
I started as a data engineer building pipelines — where you learn fast because something is always breaking. From there I trained as a data scientist, and over time grew into the role of a Solutions Architect. Today I lead the architecture for real-time data platforms and the AI systems that sit on top.
I work the seam between business intent and systems that have to run for a decade — Snowflake-native, AWS-native, Anthropic-powered, governed, observable. I write the architecture. I write the IaC. I sit in the ARB. I write the audit response, too.
I don't believe AI replaces the architect — I believe it expands the surface a single architect can design and operate. Claude reviews my pull requests. Cortex Code refactors my stored procedures. MCP servers give my agents governed access to the same systems I'm trusted with. The work becomes higher-leverage, not lower-skill.
Every system I architect today is built with both in mind: human-readable for the engineers who maintain it, machine-grounded for the agents that operate it.
Scope work to a measurable business outcome — cost reduction, latency target, audit pass, time-to-onboard — and architect backwards from there.
Row-access policies, masking, audit trails, PrivateLink, KMS — designed in, not bolted on. Compliance is a property of the platform.
Anthropic Claude, Cortex Code, RAG — applied to NL analytics, governed self-serve, migration acceleration. Token budgets per workspace, evals before production.
Pulumi for infra. GitHub Actions for SQL deploys. PR-gated cost regressions. The platform you can recreate in a weekend survives reorgs.
AI shouldn't live in a sidecar microservice nobody owns. Cortex Code and Claude run inside the same governance, observability, CI/CD perimeter.
Same architecture should behave the same at 50K and at 300K. Predictable beats clever. The 3 a.m. pager is the real architecture review.
Production GenAI for enterprises — Anthropic Claude, Snowflake Cortex Code, RAG over governed data, agentic workflows. Row-access enforced, citations grounded, evals run before promotion.
AWS-native streaming: Kinesis, Apache Flink, ECS Fargate, NLB → Snowpipe Streaming into Snowflake. Sub-second visibility, sub-300ms p99 latency. Predictable behaviour at scale.
FinOps for both compute and AI tokens. Per-tenant warehouse routing, query tagging, auto-suspend. AI workspace token caps, prompt caching, model routing (cheap-first → escalate to Claude Opus).
MCP servers giving Claude, Cursor, downstream agents governed access — Snowflake queries, audit history, design docs. Separate RAG indexes per domain with hybrid retrieval & citation grounding.
Migrating legacy SQL Server, MySQL, DB2, Oracle, Netezza, Redshift estates into Snowflake. CDC pipelines, Dynamic Tables, Streams, Tasks, SCD1/2, audit + rejection-handling frameworks.
Collapse 100+ per-customer schemas into governed shared models with row-access policies, dynamic masking, RBAC, PHI/PII protection. Tenant onboarding from weeks to hours.
Authoring Technology & Architecture Change Requests (TACR), leading external audit cycles, navigating compliance regimes (HIPAA, CJIS, SOC2). Independent second-opinion on existing designs.
Pulumi (Python) for IaC. GitHub Actions for SQL + agent deploys. PR-gated cost regressions. Docker, Jenkins, CloudWatch + Airflow observability. CI/CD that data + AI workloads share.
// 7 certifications · SnowPro Core · Azure Data Engineer · AWS · GCP · Databricks Platform Architect · Databricks Lakehouse · GitHub Professional · IBM Kafka
/01 · 2025–presentLEAD ARCHITECTReplaced a 10-VM, 25,000-line C++/MariaDB/Oracle stack with an AWS + Snowflake streaming platform. 50K devices at GA, scaling to 300K+ at ~83 TPS, sub-300ms p99 ACK, sub-90s end-to-end visibility. Authored the ARB TACR; led four external AI audit cycles to ~95% legacy parity. Pulumi IaC end-to-end.
/02 · Feb–May 2025SOLUTIONS ARCHITECTMigrated SQL Server, MySQL, Redshift to Snowflake with OpenFlow CDC, dynamic tables, dynamic masking for PHI/PII. Modernized Tableau reporting layer; CI/CD via GitHub Actions.
/03 · Apr 2025–presentSOLUTIONS ARCHITECTReusable ELT framework on Snowflake + Python on Linux. SCD1/2 merges, audit and rejection-handling frameworks, multithreaded DB2 OMS ingestion, OAuth + key-pair auth.
/04 · Feb 2025–presentAI ARCHITECTCortex Code + Snowflake Intelligence over governed data — natural-language analytics on Cortex Analyst (structured) and Cortex Search (docs), row-access enforced per tenant.
/05 · 2024DATA ENGINEERHybrid AWS/Azure data platform; Databricks + PySpark transformations; custom Kafka Connect to 15+ sources; Jenkins + Docker CI/CD; CloudWatch + Airflow observability.
/06 · 2020–2022DATA ENGINEERAzure Databricks + Data Factory ETL platforms with batch and streaming PySpark workflows. Star Performer + Excellence in Data Engineering awards.
Snowflake-native, AWS-native, Anthropic-powered. Open to staff / principal architect roles, advisory, or interesting platform problems. Replies usually within a few hours, Kansas City time.
"The best engineers don't just write code.
They architect futures worth running
for a decade — and quietly enough that
nobody has to talk about them at 3 a.m."
— Sreesanth Reddy Samala · 2026