0%
/ AI INFRASTRUCTURE   / BACKEND ARCHITECTURE   / MLOPS & VECTOR SEARCH

SAMEER
BANCHHOR

AI Infrastructure & Systems Engineer specializing in sub-15ms vectorized query execution, high-performance Go microservices, Kubernetes container orchestration, Terraform IaC, and production model serving.

Hugging Face → GitHub Profile → LinkedIn → Email Direct →
GO 1.24 REST API◆ DUCKDB 1.5 VECTOR SEARCH◆ KUBERNETES EKS & GKE◆ TERRAFORM IAC◆ VITS NEURAL AUDIO◆ DOCKER & ARGOCD◆ PROMETHEUS & GRAFANA◆ GEMINI API TRANSLITERATION◆ GO 1.24 REST API◆ DUCKDB 1.5 VECTOR SEARCH◆ KUBERNETES EKS & GKE◆ TERRAFORM IAC◆ VITS NEURAL AUDIO◆ DOCKER & ARGOCD◆ PROMETHEUS & GRAFANA◆ GEMINI API TRANSLITERATION◆
1.05M+
Voter Data Vector Indexed
DuckDB 1.5 CGO engine executing sub-15ms vectorized search
248.5K+
Batch Records Transformed
Fault-tolerant Python & Go ETL ingestion & inference pipeline
< 15ms
REST API Latency SLA
Google Cloud Run containerized Go 1.24 microservices
18M+
Neural Audio Linguistic Reach
Chhattisgarhi VITS deep neural TTS serving 18M+ native speakers
5 Targets
Parallel Multi-Output Inference
Concurrent ML prediction engine delivering real-time signals
— CORE DOMAINS

ENGINEERING CAPABILITIES

Production-proven backend architectures designed to handle intensive AI workloads and data scale.

// Backend & APIs
core.language → Go (Golang 1.24), RESTful APIs, OpenAPI 3.0
middleware → Rate Limiting (Token Bucket), Structured Logging & Tracing
— VAULT SHOWCASE

AI SYSTEMS & INFRASTRUCTURE

Selected production systems, high-throughput APIs, container orchestration, and deep neural inference pipelines.

PRODUCTION API • GO 1.24 & DUCKDB 1.5

Durg Voter Production REST API & Analytics Engine

2026 • Production Live

High-performance Go RESTful API & Embedded Analytics Engine serving 1,045,426 voter records and 1,513 polling booths with query latency strictly under 15ms. Deployed on Google Cloud Run with custom domain voter.sameerbanchhor.com.

Integrated DuckDB 1.5 vectorized analytics engine for instant electorate demography calculations.
Architected enterprise middleware: Token Bucket Rate Limiter, Request ID tracing, structured JSON logging.
Built AI transliteration engine (English ↔ Hindi via Gemini API) with <50ms fast fallback search mode.
Go 1.24 DuckDB 1.5 Google Cloud Run Gemini API Docker OpenAPI 3.0
// Key Architecture Breakdown: 1. Vectorized Analytics Engine: Direct CGO bindings to DuckDB 1.5 for instant aggregation over 1.04M records in <15ms. 2. Multi-Cloud Deployment: Primary node on Google Cloud Run with Hugging Face Space fallback mirror. 3. AI Transliteration Pipeline: Gemini API query normalization with fallback to sub-50ms fuzzy string indexing.
DEVOPS PLATFORM • KUBERNETES & TERRAFORM

Multi-Region Kubernetes (EKS/GKE) GitOps Platform with ArgoCD & Terraform

2025 • Production

Enterprise-grade container orchestration platform engineered with Terraform (IaC) and ArgoCD GitOps for zero-downtime deployments across AWS EKS and Google Cloud GKE. Features auto-scaling HPA, NGINX Ingress, Cert-Manager TLS, and unified Prometheus/Grafana metric observability.

Designed modular Terraform IaC templates automating VPC creation, IAM security roles, K8s node groups, and S3 state backends.
Configured ArgoCD GitOps workflows achieving 100% declarative cluster synchronization and instant rollback capabilities.
Kubernetes (K8s) Terraform ArgoCD Docker Helm Prometheus & Grafana
// Key Infrastructure Specs: - Single-source-of-truth Git deployment via ArgoCD syncing Helm releases directly from security-hardened repositories. - Custom Grafana dashboards tracking CPU/RAM throttling, ingress latencies, and container panic restarts.
DATA ENGINEERING • SCALABLE ML PIPELINES

DURG-EduAI — Large-Scale Data Pipeline & Multi-Task Inference Engine

2025 – 2026

High-throughput data engineering pipeline designed for processing 248,539 academic records with concurrent multi-target inference.

Built scalable ETL feature pipelines in Python processing 248.5K structured records across a 10-year dataset.
Architected 5-target parallel inference system delivering SGPA regression (R² ~ 0.996) and risk alerts in real-time.
Python ETL Pipelines XGBoost LightGBM Scikit-learn
// Key Capabilities: - Parallel multi-output engine computes 5 distinct predictions (SGPA, ATKT, Dropout Risk, Benchmarks, Alerts) per student in a single pass.
AUDIO INFRASTRUCTURE • NEURAL SYNTHESIS

Chhattisgarhi Neural Audio Synthesis Pipeline (VITS Deep TTS)

2025 – 2026

End-to-end neural audio synthesis pipeline and model serving framework for Chhattisgarhi dialect (serving 18M+ native speakers).

Architected neural synthesis pipeline from Devanagari text normalization to raw waveform audio generation.
Optimized VITS model inference pipeline for low-latency audio generation on constrained hardware.
VITS Neural Network PyTorch Model Serving Audio Processing HuggingFace Hub
// Key Highlights: - Deep VITS neural network integration with custom Devanagari phonetic processing.
— TECH STACK

ENGINEERING CAPABILITIES & TECHNICAL VAULT

Technologies and tools used to build robust, scalable production systems.

Backend & APIs
Go (Golang 1.24) RESTful APIs OpenAPI 3.0 Rate Limiting (Token Bucket) Structured Logging & Tracing
Databases & Vector Engines
DuckDB 1.5 Vectorized Search Engine SQL Optimization Low-Latency Queries Database Indexing
DevOps & Cloud Deployment
Kubernetes (K8s) Terraform (IaC) Google Cloud Run Docker Containerization HuggingFace Spaces & Hub Gemini API Git / GitHub Actions
Data Engineering & ML
Python
— BACKGROUND

ACADEMIC & CERTIFICATIONS

Educational qualifications and professional certificates.

M.Sc. in Computer Science

Hemchand Yadav University, Durg
8.21 SGPA
2024 – 2026

B.Sc. in Computer Science

Kalyan PG College, Bhilai
79%
2021 – 2023

Google Data Analytics Professional Certificate

Coursera / Google
Completed
2026

IBM Data Science Professional Certificate

Coursera / IBM
Completed
2026