██████╗ ██╗ ██╗███████╗██╗ ██╗██╗ ██╗ █████╗ ██████╗
██╔══██╗██║ ██║██╔════╝██║ ██║██║ ██╔╝██╔══██╗██╔══██╗
██████╔╝██║ ██║███████╗███████║█████╔╝███████║██████╔╝
██╔═══╝ ██║ ██║╚════██║██╔══██║██╔═██╗ ██╔══██║██╔══██╗
██║ ╚██████╔╝███████║██║ ██║██║ ██╗██║ ██║██║ ██║
╚═╝ ╚═════╝ ╚══════╝╚═╝ ╚═╝╚═╝ ╚═╝╚═╝ ╚═╝╚═╝ ╚═╝
Founding Engineer · Backend & Cloud Systems · AI/GenAI
┌─────────────────────────────────────────────────────────────────────────┐
│ $ cat impact.log │
└─────────────────────────────────────────────────────────────────────────┘
| 🚀 Metric | 📊 Result | 🛠 Method |
|---|---|---|
| Users Served |  |
Production backend shipped |
| Time to Revenue |  |
Post-launch execution |
| Infra Cost |  |
ARM migration |
| OOM Crashes |  |
Heap profiling |
| CI/CD |  |
Self-healing pipelines |
┌─────────────────────────────────────────────────────────────────────────┐
│ $ ls -la /specializations │
└─────────────────────────────────────────────────────────────────────────┘
drwxr-xr-x backend_systems/ → APIs, async queues, job processors
drwxr-xr-x cloud_infrastructure/ → AWS, Docker, K8s, Terraform
drwxr-xr-x ai_genai_engineering/ → RAG pipelines, LLMs, embeddings
drwxr-xr-x devops_sre/ → CI/CD, heap profiling, cost ops
-rw-r--r-- philosophy.txt → minimal, fault-tolerant, no drama
┌─────────────────────────────────────────────────────────────────────────┐
│ $ tech --list --filter=production-grade │
└─────────────────────────────────────────────────────────────────────────┘
Languages & Runtimes
Backend & APIs
Data & Storage
AI / GenAI Systems
Infrastructure & Cloud
Observability
┌─────────────────────────────────────────────────────────────────────────┐
│ $ tail -f /var/log/learning.log │
└─────────────────────────────────────────────────────────────────────────┘
[INFO] Hybrid search — BM25 + dense retrieval + cross-encoder reranking
[INFO] Agentic patterns — ReAct, Plan-and-Execute, LangGraph
[INFO] RAG evaluation — RAGAS, RAPTOR, Self-RAG architectures
[INFO] MLOps infra — LangFuse, experiment tracking, production evals
[INFO] pgvector — embedding storage at scale inside Postgres
[DEBUG] Reading: RAGAS, RAPTOR, ReAct, Self-RAG (12-week paper sprint)
📐 System Design notes → systemdesign.nishitdevs.xyz
┌─────────────────────────────────────────────────────────────────────────┐
│ $ cat ~/.engineering_principles │
└─────────────────────────────────────────────────────────────────────────┘
class EngineeringPhilosophy:
execution = "Ship it. Harden it. No over-engineering."
reliability = "Systems that stay up when traffic spikes."
cost = "Every dollar of infra spend is a deliberate choice."
code = "Flat structure. Direct calls. No unnecessary abstractions."
ai_usage = "Minimal, pragmatic, always questioned."┌─────────────────────────────────────────────────────────────────────────┐
│ $ git log --stat --summary │
└─────────────────────────────────────────────────────────────────────────┘
┌─────────────────────────────────────────────────────────────────────────┐
│ $ ping pushkar --open-to=US-remote-roles │
└─────────────────────────────────────────────────────────────────────────┘
Open to: Early-stage startups · AI/GenAI roles · US remote · Backend / Infra / MLOps
> Systems that survive. Code that ships. No drama.
Built with intention, not a template generator. ⚡