Google Cloud AI Builder Cup 2026
Google Cloud · Hack2skill · submitted 18 Oct 2026 · evaluation 19 Oct – 6 Nov
Project: EMBER — weather-aware fire triage
booting control plane.
Suhail Ahmed
I design and build AI agents, LLM routers & RAG systems on cloud and data platforms.
I sketch the architecture first, then build it, then measure whether it actually works. This is where I share what I’m designing, building, breaking and learning: open-source labs, hackathon builds, and notes for fellow tinkerers. 10+ years of cloud and infra under the hood.
Solutions Architect · Singapore
Now building / tinkering
- AI Builder Cup 2026: EMBER submitted, awaiting results
- llm-auto-router: held-out test split next, so the 98% gets an honest check
- Writing: first post on routing prompts to the cheapest capable model
01 / Projects
Things I’ve designed and built.
Each card has the key design decision, straight from the repo’s README.
Lab98% routing · ~27–29% cheaper
LLM auto-router on Azure AI Foundry
Routes each prompt to the cheapest model that can handle it. 98% routing accuracy on 100 labelled prompts vs 62% for keyword rules, and about 27–29% lower cost than sending everything to gpt-5.4 (measured in the repo; 98% is optimistic until a held-out split is added).
Design: An LLM classifier (gpt-5-mini) picks one of four model tiers; its JSON is validated against policy rules, and invalid output falls back to the strongest model. A 100-prompt golden set drives the tuning loop.
- Azure AI Foundry
- Python
- Evals
LabPhases 1–7
Production-grade RAG
A RAG build taken through seven phases, with a walkthrough, architecture diagram, and demo video.
Design: Hybrid BM25 + vector search (alpha 0.5, top 30), a cross-encoder reranker keeps the best 5, superseded docs are filtered out, and a CI eval gate fails any PR that drops retrieval quality.
- RAG
- Python
- Retrieval
HackathonAI Builder Cup 2026
EMBER — weather-aware fire triage
Bahasa-first AI fire commander for village brigades during Southeast Asian haze fires: ranks satellite detections by spread risk, with a volunteer app and supervisor console. Submitted; awaiting results.
Design: Agents fuse reports from many phones into one shared incident; a human supervisor approves every high-stakes call, and the WITHDRAW-upwind safety rule is deterministic code, not model output.
- Gemini Live
- ADK
- Cloud Run
- Firestore
HackathonSubmitted
Fraud policy copilot
Snowflake CoCo CLI fraud case copilot: transaction in, policy match, risk score with citations. Risk / Fraud / Regulatory track.
Design: Streamlit in Snowflake: transaction id → customer context → rule matches → risk score with policy citations, then a case and a native Snowflake alert trail.
- Snowflake
- CoCo CLI
- Python
LabCI gate
Foundry eval gate
A CI eval gate that checks agent traces for faithfulness, tool success, and PII leaks before a change ships.
Design: Deterministic, fixture-first checks on JSON traces that exit non-zero, so GitHub Actions or Azure DevOps can block the merge.
- Evals
- CI/CD
- Agents
LabHuman-approved
ADO pipeline doctor
Turns failed Azure Pipelines and GitHub Actions logs into structured RCA JSON, with human approval required before any fix.
Design: Logs become structured RCA JSON with an explicit human_approval_required flag, so it’s safe to wire into an agent.
- Azure DevOps
- GitHub Actions
- LLM
LabMCP toolkit
MCP-style ops tools
iac-guard-mcp (Terraform/Bicep diff review), azure-ops-mcp (Key Vault compare, ADO diagnose, dry-run + audit), lakehouse-insights-skill, and kc-mcp.
Design: iac-guard uses Checkov when available and falls back to built-in regex policies, returning structured findings JSON that agents can cite.
- MCP
- IaC
- Azure
LabIaC
Infrastructure as code
Hub-and-spoke Terraform modules for Azure networking, Python Key Vault automation, and multi-CI demos across GitHub Actions, Azure DevOps, and GitLab.
Design: Separate hub and spoke Terraform modules plus an example that wires them together: a foundation to extend, not a full landing zone.
- Terraform
- Azure
- Python
- CI/CD
More on GitHub.
02 / Hackathons
Where I build under a deadline.
Snowflake CoCo CLI Hackathon — GCC Edition
Snowflake · Hack2skill · Sep–Oct 2026
Project: Fraud policy copilot
Nebius x NVIDIA Global AI Hackathon
Nebius x NVIDIA · Devpost · Oct 2026
Build with AI: Code for Communities — Second Edition
Hack2skill · Sep 2026
HACKATHON X FIT-FEST 2026
FIT-FEST · Hack2skill · Sep 2026
Databricks GenAI Hackathon 2024 — team REV-Engers
Time-boxed team GenAI hackathon · 2024
RAG experiments on Databricks: table setup, CSV/PDF document ingestion, a RAG chain agent model, and a model GUI. Notebooks on GitHub
03 / Writing
Notes from the build, for fellow builders.
Routing prompts to the cheapest capable model
How the LLM auto-router beat keyword rules. Project: LLM auto-router
Posts on LinkedIn; each links back to its project card here.
04 / About
Hi, I’m Suhail. I design AI systems, then build them until they actually run.
In 30 seconds
I’m a solutions architect who still writes the code. I design and build AI agents, LLM routers and RAG systems on cloud and data platforms, after 10+ years in cloud and infra on Azure and AWS. These days it’s agents, evals and making models cheaper to run. Some of it runs in production at my day job; the rest lives on GitHub.
In 2 minutes
I started out automating VMs by the thousand, then moved into cloud migrations, landing zones and Terraform pipelines, then DevOps: Kubernetes, observability and lots of IaC.
Now I design and build the AI layer. At my day job that means an Azure AI Foundry gateway that governs how AI gets used, and AI agents on Databricks, Snowflake and Sigma that save people from manual analysis. I care about evals and operational readiness more than demos.
On my own time I design and build things to learn, starting from an architecture diagram: an LLM auto-router that picks the cheapest model that can do the job, a production-grade RAG, MCP tools for ops, and hackathon builds like EMBER. If it breaks, I write down why.
TL;DR for builders
- Designs & builds: AI agents, LLM routing, RAG, evals, MCP tools
- How: architecture first, then code, then evals to check it works
- Runs on: Azure AI Foundry, Databricks, Snowflake, Google Cloud / Gemini
- Under the hood: 10+ years of Azure, AWS, Terraform, Kubernetes and CI/CD
- Right now: waiting on AI Builder Cup results; adding a held-out eval to the LLM auto-router
- Projects · GitHub · Resume (PDF)
10+years in infrastructure, cloud & data
9certifications
20public GitHub repos
Stack I play with: tap one to see where it shows up
Pick a tool to light up the projects, hackathons and production work that use it.
05 / Running in production
At my day job, this stuff runs for real.
Azure AI Foundry AI gateway
An Azure AI Foundry gateway I designed and implemented to manage, monitor and govern AI solutions: one place for model access, oversight and control.
AI agents on Databricks, Snowflake & Sigma
AI agents I designed on the data platform that cut manual analysis for ops and business teams, plus AI for analytics and operational insights.
Readiness & responsible tooling
Evals, testing and operational readiness for AI apps; trying out Cursor and MCP with security and configuration first.
Scope only. Internal details and numbers stay at work.
06 / Certifications
Azure Solutions Architect Expert
DevOps Engineer Expert
AWS Solutions Architect
Terraform Associate 004
Databricks Data Engineer
SnowPro Core
AI-103 Azure AI Apps and Agents
Azure AI Fundamentals
OCI Architect Associate
07 / Career path
Ten years from VMs to AI agents Show the full career path
Solutions Architect — Cloud, Data & AI
Revantage Global · Singapore
Responsibilities
- Enable AI across Databricks, Snowflake, and Sigma for analytics and operational insights.
- Design AI agents that cut manual analysis for Ops and business teams.
- Implement Azure AI Foundry as a gateway to manage, monitor, and govern AI solutions.
- Evaluate Cursor and MCP with a focus on security, configuration, and responsible use.
- Support AI application build-out, evals, testing, and operational readiness.
Senior DevOps Engineer
Revantage Global · Singapore
Responsibilities
- Design enterprise Azure and AWS architectures, including Hub-and-Spoke landing zones.
- Build IaC and CI/CD with Terraform, Bicep, Azure DevOps, and GitLab.
- Run Kubernetes, Docker, and ECS/Fargate for availability and cost.
- Implement Prometheus, Grafana, observability, and cloud security controls.
Cloud / DevOps Consultant
Accenture Singapore
Responsibilities
- Migrate complex legacy applications to Azure and AWS with resilient, cost-aware designs.
- Build secure landing zones with Terraform and GitLab CI/CD to cyber standards.
- Deliver cloud security and compliance projects with guardrails on Azure workloads.
Cloud / DevOps Engineer
Accenture India · Bangalore
Responsibilities
- Deploy IaaS on AWS and Azure using Terraform and GitLab CI/CD.
- Execute application migrations and performance optimizations.
- Investigate incidents and work remediations with compliance and security teams.
Technical Lead
Cognizant · Bangalore
Responsibilities
- Automate cloud management with vRealize Automation and Orchestrator across 1000+ VMs.
- Integrate vRO with CouchDB and manage ESXi, HA, DRS, and migrations.
- Run Agile delivery and Bitbucket workflows; training that lifted team skills.
08 / Contact