HC
HC
GENAI & BACKEND ENGINEER

HARSH CHAURASIAHARSH CHAURASIA

Building high-performance microservices, RAG architectures, and agentic AI models. Specializing in Golang and Python backends operating at production scale.

SEE MY WORKLET'S TALK
// 01. CORE PROFILE

WHO I AM & WHAT I STAND FOR

I operate at the intersection of reliable backend infrastructure and practical artificial intelligence. My work centers on building production-grade distributed systems, RAG pipelines, and agentic platforms in Golang and Python.

I believe backends should be fast, stateless, and containerized. Whether designing gRPC communication layers, scaling SQL & Qdrant query databases, or deploying with Docker Compose, I write systems designed to survive high concurrent loads.

Instead of standard boilerplate, I focus on coding developer-oriented tools and autonomous scrapers. I don't just deploy systems; I engineer software designed to act intelligently.

Currently Building
davincer
AI Video Editor
Apply-AI
Job Form-Fill automation
Invisor
WebSocket Finance streams
* Thread logs auto-synced with repository manifests.
// 02. SYSTEM COEFFICIENTS

TECHNICAL MATRIX

Languages

Golang (Primary)PythonTypeScriptJavaScript

AI & GenAI

RAG PipelinesAgentic AI SystemsLLM OrchestrationPrompt EngineeringVector Search (Qdrant)Reranking ModelsStructured OutputsSTT/TTS PipelinesLLM GuardrailsCLEAN Prompting

Backend & Systems

MicroservicesREST APIsgRPC ProtocolsWebSocketsPostgreSQLDAO/Service/HandlerDocker & ComposeGin / Echo (Go)FastAPI (Python)LangChainNode.jsCobra CLI Tooling

Cloud & DevOps

GCP (Cloud Run / Storage)GitHub ActionsCI/CD PipelinesLinux CLI / BashNginx Reverse ProxyPM2 Process ControllerSSL CertbotKubernetes (Basics)
// 03. LOG TIMELINE

EXPERIENCE LOG

Software Engineer

BlockSynergy Technology Limited
May 2025 – Mar 2026

Blockchain-backed lending platform — microservices, Agentic AI, and RAG at production scale.

>Designed and owned a microservices backend in Golang with PostgreSQL, Docker, and 300+ REST APIs powering end-to-end digital lending workflows.
>Built an Agentic AI lending assistant (LLM + RAG) for loan retrieval, EMI insights, contextual recommendations, and support ticket resolution — integrated directly into business workflows.
>Architected RAG pipelines with vector retrieval and reranking, measurably improving response relevance for financial Q&A.
>Developed CLI tooling in Go (Cobra) for blockchain infrastructure provisioning and private network generation.
>Integrated Razorpay payment gateway, notification systems, and event-driven workflows across microservices.
>Led a cross-functional team of 5 engineers across backend, AI, and integrations, coordinating architecture decisions and delivery timelines.
>Containerized all services with Docker Compose; contributed to GCP-based cloud deployments.
>Represented the company at UP International Trade Show (UPITS 2025).

Co-Founder & Backend Lead

Ivora AI
Dec 2023 – Feb 2025

Early-stage AI startup building voice based AI agents.

Backend Developer (Golang)

Freelance
Oct 2022 – Nov 2023

Developed production Golang applications using layered DAO/Service/Handler architecture for multiple clients.

// 04. SHIPPED APPLICATIONS

FEATURED PROJECTS

davincer

AI-powered video editing platform with chat-driven automation.

>>Enabled chat-driven automated video edits, letting users describe timeline alterations in natural language.
>>Engineered a timeline generation engine with media importing and beat synchronization.
>>Architected scalable frontend interface with strict TypeScript design patterns.
Next.jsReact 19TypeScriptNode.jsTailwindCSS
IMPACT:10x faster video editing timeline compilation
TERMINAL // davincer.sh
$ npm run dev --davincer
> [Core] Analyzing video assets...
> [AI] Prompt: 'Cut timeline at high beat changes'
> [Sync] Matched 18 audio transients with video cuts
> [Timeline] Render completed in 1.2 seconds.
$

Apply-AI

Autonomous job application platform with scraping and form-fill automation.

>>Automated scraping and form submission workflows across LinkedIn, Indeed, and Naukri.
>>Built custom AI orchestration layer that handles model rotation and browser failovers.
>>Utilized Playwright and Crawlee for stealth browser automation and reliable data fetching.
GolangNode.jsReactPostgreSQLDockerPlaywrightCrawlee
IMPACT:95% form-fill accuracy and automatic cover letters
TERMINAL // Apply-AI.sh
$ go run apply-ai.go
> [Orchestration] Launching Apply-AI cluster...
> [Playwright] Bypassing Cloudflare protection...
> [Form] Populating application data fields...
> [AI] Custom cover letter generated & submitted.
$

Invisor

AI financial advisory backend serving real-time market insights.

>>Ingested real-time market ticker feeds and computed technical indicators.
>>Executed low-latency sentiment analysis on streaming news using Python NLP models.
>>Surfaced calculated trading insights via optimized WebSocket broadcast channels.
GolangPythonPostgreSQLWebSocketsFastAPI
IMPACT:Sub-50ms end-to-end WebSocket dispatch latency
TERMINAL // Invisor.sh
$ python main.py --stream
> [Socket] Connected to live equity exchange...
> [Stream] Ingesting ticker feeds for: AAPL, GOOG, NVDA
> [NLP] Sentiment polarity: +0.89 (Bullish)
> [Websocket] Dispatched alert to 240 active clients.
$
// 05. REACH OUT

ESTABLISH CONNECTION

SYSTEM DIAGNOSTICS
Availability Stream
LIVE
HOST NODE:harsh_core // remote
NET_STABILIZER:42ms // stable
GATEWAY SECURE:TLS_1.3_ACTIVE
HANDSHAKE PROTOCOL:ACCEPTING SIGNALS
COMMUNICATION MATRIX:
FULL-TIME
Enterprise dev
FREELANCE
Integrations
COLLABS
GenAI hacks
TERMINAL // TRANSMIT_PAYLOAD.SH
"";
"";
"";
EMAIL
[email protected]
GITHUB
github.com/harshc9
LINKEDIN
linkedin.com/in/harsh-chaurasia
PORTFOLIO // HARSH CHAURASIA // VERSION 2.0
BUILT WITH NEXT.JS 14, TAILWIND CSS & FRAMER MOTION