Skip to main content
HB

All Projects

Filter by category

Full-Stack
Freelance client -- live in production

Markaz Idara Tul Islah — An ERP for an Institute That Ran on Paper

An institute running on paper registers, digitalized end-to-end -- bilingual Urdu/English, live with 200+ users.

200+Active Users — Live in Production
Full-StackDatabaseSecurityBackend & APIs
Next.jsExpressPostgreSQLRedis+5
View Case Study Demo
Platform & Cloud
Built at QisstPay (2022–2024)

QisstPay — Hybrid Cloud Infrastructure for a BNPL Fintech

Hybrid AWS + GCP infrastructure for a BNPL fintech -- 17 microservices, PCI DSS scope, bank VPN integrations.

1M+Daily Transactions on Infrastructure I Operated
AWSGCPSecurityDatabase+3
AWS ECSGCP Cloud RunTerraformMySQL+10
View Case Study Demo
Platform & Cloud
Freelance client

Production EKS Platform with SRE Observability

An EKS platform with real SRE discipline -- SLOs, Karpenter, and 99.9% measured over six months.

99.9%Uptime Over 6 Months
AWSKubernetesCI/CDObservability & SRE+1
AWS EKSKubernetesKarpenterHelm+8
View Case Study
Platform & Cloud
Freelance client -- Upwork

Bare-Metal K3s Cluster with PostgreSQL HA

Six hand-configured servers to a reproducible K3s + Postgres HA stack -- full rebuild in under 30 minutes.

<30minFull Stack Bootstrap
KubernetesDatabaseSecurityAutomation+1
K3sAnsiblePostgreSQLetcd+8
View Case Study
Platform & Cloud
Freelance client

High-Availability Booking Platform — AWS ECS on EC2

Fargate pricing said $8k/month. ECS on EC2 with Reserved Instances said $3.8k. High-availability booking at scale for less.

99.95%Uptime · 9 Months
AWSTerraform & IaCCI/CDDatabase+3
AWS ECSEC2ECRTerraform+10
View Case Study Demo
Platform & Cloud
Freelance client

AI-Powered Legal Document Analysis API

Productizing a legal RAG pipeline: auth, rate limits, and 99.95% uptime over six months.

99.95%API Uptime Over 6 Months
GenAIBackend & APIsDatabaseSecurity+1
FastAPIPythonSupabasePostgreSQL+6
View Case Study
Platform & Cloud
Freelance client

GDPR and SOC 2 Compliant Data Pipeline on AWS

GDPR erasure across 12 systems in 72 hours, with Step Functions execution logs as audit evidence.

72hrsRight-to-Erasure SLA Across 12 Systems
AWSSecurityDatabaseTerraform & IaC+2
AWS GlueAWS KMSS3AWS Macie+8
View Case Study Demo
Platform & Cloud
Freelance client

Multi-Cluster GitOps Platform with ArgoCD

14 services, 3 clusters, zero drift -- 120 deploys/week through ArgoCD with auto-rollback.

120/weekAutomated Deployments with Zero Drift
KubernetesCI/CDSecurityObservability & SRE+1
ArgoCDKubernetesKustomizeHelm+7
View Case Study
MLOps & Model Serving
Freelance client

Handwritten Document OCR Platform

OCR pipeline for degraded handwritten manuscripts -- preprocessing did the heavy lifting, not the model.

99.4%Character-Level Accuracy After Preprocessing
GCPMLOpsGenAIModel Deployment+2
PythonFastAPIGCP Cloud RunGoogle Vertex AI+6
View Case Study
MLOps & Model Serving
Freelance client

Serving an 18GB ML Model on Cloud Run

An 18GB transformer on serverless: cold start cut from 2+ minutes to 8 seconds with gcsfuse streaming.

8sCold Start (from 2 min)
GCPMLOpsModel DeploymentServerless+1
PythonGCP Cloud RungcsfuseDocker+5
View Case Study Demo
MLOps & Model Serving
Freelance client -- Upwork

SageMaker ML Model Deployment with Auto-Scaling

A model that worked on laptops and died in prod -- rebuilt as an autoscaled SageMaker endpoint, <200ms P99.

<200msP99 Inference Latency
AWSMLOpsModel DeploymentBackend & APIs+1
AWS SageMakerAPI GatewayLambdaS3+5
View Case Study
MLOps & Model Serving
Freelance client

Custom LLM Serving on Bare Metal GPUs with llama.cpp

200 attorneys, $18k/month in API calls -- moved to self-hosted A100s at 82% lower cost.

82%Cost Reduction vs Cloud GPUs
MLOpsModel DeploymentGenAIObservability & SRE+1
llama.cppNVIDIA A100CUDA 12.2GGUF+8
View Case Study
RAG & AI Agents
Freelance client -- Upwork

Production RAG Pipeline with pgvector

Replaced an $800/mo Pinecone instance with pgvector -- measured on a domain golden set, not vibes.

95%Recall@5 on a Domain Golden Set
RAGGenAIDatabaseBackend & APIs+1
PythonFastAPISupabasePostgreSQL+4
View Case Study
RAG & AI Agents
Freelance client

Multi-Stage Legal RAG with Azure AI Search

Legal RAG that respects document structure: 5 retrieval layers over 3,145 docs, each layer fixing a real failure.

3,145 Docs5-Layer Retrieval Pipeline
AzureRAGGenAIAI Agents+3
PythonFastAPINext.jsAzure AI Search+8
View Case Study
RAG & AI Agents
Freelance client

MCP Server for ERP Integration

214 undocumented ERP tables turned into 42 curated MCP tools an agent can actually use safely.

42 Tools90% of Sampled User Questions Answerable via Tools
AI AgentsGenAIDatabaseBackend & APIs+1
MCP ProtocolPythonFastAPIPostgreSQL+6
View Case Study
Automation
Freelance client -- Upwork

UK Planning Intelligence Scraper and CRM

A JS-rendered council portal with session pagination, scraped responsibly -- zero missed applications.

0 MissedApplications Across 3 Clean Runs
AutomationBackend & APIsDatabaseSecurity
PythonFastAPISupabasePostgreSQL+8
View Case Study