Hi, I'm
SahilsinghKhalsa.
United States
I build AI-powered software that turns messy real-world workflows into reliable, usable products. My work spans full-stack engineering, backend systems, cloud infrastructure, DevOps, RAG pipelines, and agentic AI applications built for real users, not demos.
About Me
A quick look at who I am and what I do.
Two years ago, I came to the United States from India with uncertainty, pressure, and a clear goal: become someone who can build real things, not just talk about them.
During my Master's at Illinois Tech, I started seeing software as more than assignments. It became a way to solve messy problems and build systems people depend on. Since then, I've gained hands-on experience at Find Me LLC, Gabriel AI, and The Special Character, working across full-stack engineering, backend systems, cloud workflows, DevOps, and applied AI.
What drives me is turning unclear problems into usable software. I like building reliable APIs, secure systems, and intelligent applications that survive beyond the demo stage. I'm still early in my career, but building with clear direction: learn fast, take ownership, and ship software that actually matters.
Backend & Systems
FastAPI · Node.js · Microservices · REST APIs
Production APIs and multi-service backends built for real traffic.
AI Engineering
LangGraph · RAG Pipelines · OpenAI · LLMs
End-to-end AI pipelines from retrieval to multi-agent orchestration.
Cloud & Infrastructure
AWS · PostgreSQL · MongoDB · Docker
Cloud-native deployments with CI/CD, observability, and zero downtime.
Education
Illinois Institute of Technology
Aug 2024 – May 2026Master of Science, Computer Science
Chicago, IL · GPA: 3.7 / 4.0
Vidush Somany Institute of Technology
Aug 2020 – May 2024Bachelor of Science, Computer Engineering
India · GPA: 4.0 / 4.0
Journey & Timeline
Professional milestones and academic background, all in one view.
Find Me LLC
Software Development Engineer
- Scaled FyndMe.net to support 9 portfolio types and 4 template tiers by architecting a 3-service microservices platform across Next.js, Node.js/Express, and FastAPI.
- Collaborated with product and engineering teams to ship 11+ public and protected workflows by translating portfolio requirements into React/Next.js pages, API routing, and JWT-secured client-to-backend communication.
- Unified 9 portfolio schemas into a single MongoDB model using Mongoose discriminators, with per-type validation, visibility controls, and soft-delete support.
- Developed a containerized AI-powered FastAPI image-processing service using U²Net, MediaPipe, and OpenCV.
Find Me LLC
Software Development Engineer
- Scaled FyndMe.net to support 9 portfolio types and 4 template tiers by architecting a 3-service microservices platform across Next.js, Node.js/Express, and FastAPI.
- Collaborated with product and engineering teams to ship 11+ public and protected workflows by translating portfolio requirements into React/Next.js pages, API routing, and JWT-secured client-to-backend communication.
- Unified 9 portfolio schemas into a single MongoDB model using Mongoose discriminators, with per-type validation, visibility controls, and soft-delete support.
- Developed a containerized AI-powered FastAPI image-processing service using U²Net, MediaPipe, and OpenCV.
Gabriel AI
Backend Engineer
Gabriel AI
Backend Engineer
The Build Fellowship
Developer Build Student Consultant
The Build Fellowship
Developer Build Student Consultant
The Special Character
Software Engineer
The Special Character
Software Engineer
Featured Projects
A selection of projects built across AI pipelines, backend systems, and full-stack SaaS.
LedgerLens: Multimodal Financial-Filing Intelligence Engine
Multimodal RAG engine for SEC financial filings that retrieves over rendered page images via ColPali v1.2, eliminating the table structure loss that breaks text-only pipelines. A 6-node LangGraph agent cross-checks every extracted figure against SEC XBRL ground truth before it ships, making hallucinated numbers architecturally impossible. Single-fact queries bypass retrieval entirely via a planner fast path, answering directly from the XBRL database in under 1 second. Includes earnings call ASR via Whisper, XBRL trend forecasting via OLS regression, and per-query cost tracking across all nodes. Achieved 83.3% numeric EM and 0% hallucination on NVDA and MSFT 10-Ks.
SIGMA: Multi-Agent AI Trading System
Full-stack autonomous trading system that monitors 20 stocks in real time, detects 9 technical signals, and routes them through a LangGraph ReAct agent that reasons across 6 research tools before deciding to trade. Every decision passes an 8-rule deterministic risk agent and streams live to a 7-page React dashboard via WebSocket. Built with a 6-call budget cap (~$0.02 per trade), 20 passing tests, Docker deployment, and GitHub Actions CI/CD.
Outreach Revival: Community Health Platform
AI-triaged volunteer dispatch platform connecting elderly and homebound residents with vetted volunteers from faith-based organizations. An 8-agent Claude network over 5 MCP servers handles intake, matching, SMS dispatch, FHIR R4 referrals, and daily risk scans. Volunteers are ranked by a 6-signal weighted formula covering distance, schedule, skills, rating, language, and relationship history.
ARIA: Adherence Risk Intelligence Agent
Full-stack clinical intelligence platform for hypertension management. Ingests patient EHR data via FHIR R4, runs a three-layer nightly AI pipeline (deterministic rule engine, numeric risk scorer, and LLM clinical briefing) and delivers a structured pre-visit briefing to clinicians at 7:30 AM on appointment days. Features patient-adaptive BP thresholds, drug interaction detection, an AI chatbot for patient Q&A, a patient PWA for home readings, and 583 passing tests.
My Skills
Technologies and tools I work with across backend, cloud, and AI.
Dots indicate proficiency: ●●●●● expert · ●●●●○ advanced · ●●●○○ proficient
Licenses & Certifications
Verified credentials across AI, cloud, and data engineering.
Introduction to Agent Skills
Anthropic
Hands-On AI: Build a RAG Model from Scratch with Open Source
AWS Generative BI with Amazon Q in Quicksight
Amazon Web Services
Amazon S3 Performance Optimization
Amazon Web Services
Python for Data Science
NPTEL
Artificial Intelligence Virtual Experience Program
Forage
Get In Touch
Have a project or opportunity? I reply within 24 hours.
Open to full-time SDE and AI engineering roles. Also happy to collaborate on freelance projects or research.