┌─────────────────────────────────────────────────────────────────────────────────────────────┐
│ │
│ ███████╗██████╗ ██╗███╗ ██╗ ██╗ ██████╗ ██╗ ██╗ ██████╗ █████╗ ███████╗ │
│ ██╔════╝██╔══██╗██║████╗ ██║ ██║██╔═══██╗╚██╗ ██╔╝ ██╔══██╗██╔══██╗██╔════╝ │
│ ███████╗██████╔╝██║██╔██╗ ██║ ██║██║ ██║ ╚████╔╝ ██║ ██║███████║███████╗ │
│ ╚════██║██╔══██╗██║██║╚██╗██║ ██ ██║██║ ██║ ╚██╔╝ ██║ ██║██╔══██║╚════██║ │
│ ███████║██║ ██║██║██║ ╚████║ ╚█████╔╝╚██████╔╝ ██║ ██████╔╝██║ ██║███████║ │
│ ╚══════╝╚═╝ ╚═╝╚═╝╚═╝ ╚═══╝ ╚════╝ ╚═════╝ ╚═╝ ╚═════╝ ╚═╝ ╚═╝╚══════╝ │
│ │
│ Low Latency Systems · Agentic AI · Quant Research · SDE │
│ [ IIT Kharagpur ] │
└─────────────────────────────────────────────────────────────────────────────────────────────┘
uptime: always load: high status: building
srinjoy@iitkgp:~$ cat about.txt
NAME : Srinjoy Das
HANDLE : SRINJOY59
NODE : IIT Kharagpur
ROLE : Systems Engineer · ML/AI Engineer · Quant Dev
FOCUS : Making software run faster than it has any right to
CREED : If it isn't measured in nanoseconds, is it even fast?
srinjoy@iitkgp:~$ ls -la domains/
drwxr-xr-x ⚡ low-latency-systems/
drwxr-xr-x 🤖 agentic-ai/
drwxr-xr-x 📊 quant-finance/
drwxr-xr-x 🏗️ software-engineering/
drwxr-xr-x 🧠 machine-learning/
srinjoy@iitkgp:~$ cat domains/low-latency-systems/README
── Low Latency & Systems Engineering ────────────────────────────────────
Language C++ (17 / 20 / 23) — primary weapon of choice
Concurrency lock-free structures, wait-free queues, atomics
Memory cache-line alignment, huge pages, memory pools
CPU SIMD/AVX intrinsics, CPU pinning, NUMA topology
Network kernel bypass (DPDK), zero-copy I/O, RDMA
Profiling perf, VTune, Valgrind, flamegraphs
Target sub-microsecond execution paths
srinjoy@iitkgp:~$ cat domains/agentic-ai/README
── Agentic AI & LLM Systems ─────────────────────────────────────────────
Frameworks LangGraph, CrewAI, custom orchestration layers
LLMs fine-tuning (LoRA/PEFT), RAG pipelines, tool use
Inference model quantization, batching, latency optimization
Agents multi-agent planning, reflection loops, memory
Libraries PyTorch · HuggingFace · LangChain · vLLM
srinjoy@iitkgp:~$ cat domains/quant-finance/README
── Quantitative Finance ─────────────────────────────────────────────────
Strategies statistical arbitrage, mean reversion, momentum
Derivatives options pricing — Black-Scholes, Heston model
Risk VaR, CVaR, Greeks, portfolio optimization
Infrastructure real-time market data feeds, order management
Backtesting event-driven simulation frameworks
Data tick data processing, feature engineering
srinjoy@iitkgp:~$ cat domains/software-engineering/README
── Software Engineering ─────────────────────────────────────────────────
Backend Django · FastAPI · Node.js · REST · GraphQL
Frontend React · TypeScript · Tailwind CSS
Databases PostgreSQL · Redis · MongoDB · SQLite
Infra Docker · Linux · Kafka · CI/CD pipelines
Design distributed systems · system design · microservices
srinjoy@iitkgp:~$ uname -a # tech stack
Languages
AI / ML
Web & Backend
Infra & Tools
srinjoy@iitkgp:~$ ps -ef --sort=-%cpu | head -8 # what's running
UID PROCESS STATE PRIORITY
──────────────────────────────────────────────────────────────────
srinjoy low_latency_cpp_optimization running RT
srinjoy agentic_ai_framework running high
srinjoy hft_strategy_backtester running high
srinjoy competitive_programming running normal
srinjoy ml_model_experiments running normal
srinjoy full_stack_project sleeping low
srinjoy@iitkgp:~$ ping linkedin.com -c 1
PING linkedin.com — 1 packet transmitted
Interested in:
→ Low latency / HFT infrastructure challenges
→ Agentic AI & LLM-powered systems
→ Quantitative research & algo trading
→ Hard engineering problems
Reply from: linkedin.com/in/srinjoy-das-174648202
RTT: fast
"Premature optimization is the root of all evil — but late optimization is just embarrassing."