Hi, I'm Jishnu
Janardanan
Software Engineer | B.S. in Computer Engineering from Illinois Institute of Technology
Building scalable applications with React and AWS. Passionate about clean code and delivering great user experiences.
About Me
I am a Computer Engineering student from Illinois Institute of Technology, graduated in May 2026. I like building things that work well and look good.
When I am not coding or designing systems, you will find me watching cricket, playing basketball, or checking out new sneaker drops.
Education
B.S. Computer Engineering
Illinois Institute of Technology
May 2026
Technical Skills
AI/ML & LLM
LangChainLangGraph
LangSmith
NeMo Guardrails
Hugging Face
RAGAS
FAISS
Model Context Protocol (MCP)
Anthropic API
OpenAI API
Google Gemini APINVIDIA NIMLanguages
Python
Java
JavaScript
C++
Dart
Bash
SQL
HTML5
CSS3Frameworks & Libraries
React
Vite
FastAPI
Flutter
Node.js
Spring Boot
JUnitOpenMP
Vue 3
pandasDatabases
PostgreSQL
MySQL
DynamoDB
Redis
FirebaseCloud & DevOps
AWS (Lambda, EC2, API Gateway, S3, CloudFront, SES, EventBridge, IAM, Bedrock)
GCP (YouTube API, Firebase Auth)
Docker
Git
GitHub Actions
Bazel
Linux/UbuntuMQTT
Terraform
Kubernetes
Prometheus
GrafanaDistributed Systems & System Design
EDA & VLSI Tools
AI/ML & Performance
AI-Assisted Development
Experience
Cloud Engineer (Project Manager)
Changing The Present
New York, NY · Remote
June 2026 - Present
- • Audited a 23-table PostgreSQL database with Python (pandas) and pgAdmin, uncovering that 31% of applicants were silently dropped and 26% of tables were dead.
- • Moved the Postgres superuser password out of plain-text config into AWS Secrets Manager and gave each service its own read-only role, with zero downtime during the cutover.
- • Designed a BCNF schema and a re-runnable ETL sync (Python, AWS Lambda, EventBridge, VPC), cutting data lag from 5 months to 1 hour (99.9%).
- • Led the team through written briefs and clear task delegation to ship a Google Sheets-to-Postgres ETL pipeline.
Backend Developer Intern
FoodCLUB
London, United Kingdom
Jan - May 2024
- • Held authentication and data APIs at 99% uptime by building and maintaining Node.js services on AWS EC2.
- • Cut response latency 30% on the busiest endpoints by adding Redis caching.
Hardware Research
IIT Motorsports (FSAE)
Chicago, IL
Jan - May 2024
- • Researched hardware architecture for quad inverter in Formula SAE electric vehicle
- • Contributed to power switch selection and component specifications
Projects
SneakersForLess
LivePrice comparison engine ranking sneaker quotes across 25+ retailers. Firebase-authenticated serverless backend with automated price alerts via EventBridge and SES.
slice – AI & Cloud Cost Gateway
LiveSelf-hosted LLM gateway that routes, caches, and caps AI spend across Anthropic, OpenAI, Gemini, and NVIDIA NIM behind one endpoint. A LangGraph router with a LoRA fine-tuned Qwen2.5-0.5B classifier (92% routing accuracy) cut a 50-prompt workload's cost 45%, with a RAGAS LLM-as-a-judge confirming 0.89 relevancy. NeMo Guardrails, fail-open Redis/Postgres paths, 608 unit tests, Terraform to arm64 EC2 with Prometheus/Grafana, and a 7-tool MCP server for Claude Code.
Old Is Gold
LiveFitness platform for users 55+ with personalized workout plans, nutrition tracking, and sub-200ms dashboard loads via single-partition DynamoDB design. JWT auth with per-user rate limiting.
Low-Latency Trading Exchange (C++)
Low-latency trading exchange built from scratch in modern C++ (C++26, Bazel). Matching engine with price-time priority, a pre-allocated memory pool, a lock-free MPSC queue, and a live market data feed with a client-side book rebuilder. Verified with 113 tests; a live run streamed 111,842 messages to 4 clients that each rebuilt an 8,235-order book from the feed alone with zero drops.
IoT Smart Device Hub
Spring Boot device management backend syncing state in real time across HTTP and MQTT using pub/sub and the observer pattern. 94% test coverage with 30+ JUnit tests.
Neural Network Inference Engine
C++ engine running MLP and CNN models end-to-end. 97.5% MNIST accuracy, 8% cache hit rate gain from NCHW/NHWC tensor layout optimization, parallelized with OpenMP.
ThinkFast Embedded Trivia Game
Raspberry Pi + Arduino trivia system over UART serial with sub-100ms LED feedback. Co-authored paper under review at IEEE EIT 2026.
CTA Database System
Relational database with 6-8 entities, Python CRUD app with complex SQL queries, window functions, and OLAP analytics.
32-bit Pipelined CPU
Custom RISC processor based on MIPS ISA with modular ALU and pipeline synchronization.
Chatbot using AWS Bedrock
Conversational chatbot built on AWS Bedrock foundation models, with a serverless Lambda backend.
EasyNN Matrix Operations
Matrix and tensor operations in C++ for the EasyNN framework with automated validation.
Certifications
NVIDIA Certified Professional: Agentic AI
NVIDIAProfessional certification validating the ability to architect, develop, deploy, and govern advanced agentic AI systems, spanning agent architecture, cognition and memory, multi-agent orchestration, evaluation and tuning, and deployment at scale on NVIDIA platforms.
Verify Credential →C++ Systems for Finance: Build a Trading Exchange from Scratch
Open Avenues Foundation — The Build Fellowship
August 2026
8-week Build Project led by Yiwen Victor Song. Built a low-latency trading exchange from scratch in modern C++: price-time priority matching engine, pre-allocated memory pool, lock-free MPSC queue, and a live market data feed with client-side book rebuilding.
Building RAG Agents with LLMs
NVIDIA Deep Learning Institute
Hands-on certification in retrieval-augmented generation: building LLM agents with vector stores, embedding-based retrieval, document ingestion pipelines, and multi-turn agent state management.
Building Agentic AI Applications with Large Language Models
NVIDIA Deep Learning Institute
Hands-on certification in agent architectures: structured outputs and function calling, retrieval and knowledge graphs, and multi-agent orchestration with LangGraph and NVIDIA NIM for long-horizon reasoning and real-time operation at scale.
Claude 101
Anthropic Academy
Foundational certification in AI-assisted workflows: prompt engineering and working effectively across Claude's feature set (Projects, Artifacts, Agent Skills, Connectors, Research).
Claude Code 101
Anthropic Academy
Foundational certification in agentic coding with Claude Code: driving development from the terminal, delegating multi-step engineering tasks, and integrating AI-native workflows into a real codebase.
Introduction to Model Context Protocol
Anthropic Academy
Certification in the Model Context Protocol (MCP): connecting LLMs to external tools and data sources through standardized servers, enabling agents to access live context and act across systems.
AI Fluency: Framework & Foundations
Anthropic Academy
Framework for effective and responsible AI collaboration, covering task delegation, prompt engineering, critically evaluating AI outputs, and diligent, ethical use across the 4D model.
Get in Touch
jishnujanardanan2003@gmail.com
Phone
+1 (312)-852-0631
Location
Chicago, IL, U.S.A (Authorized to Work)