AI Engineer  ·  AmlgoLabs  ·  Gurugram

Kunal
Aggrawal

Building production AI backends — LangGraph multi-agent systems, enterprise FastAPI services, and AWS infrastructure automation.

View Experience Get in touch
264K+ jobs evaluated in one run
~3K invoices processed daily
~80% Text-to-SQL accuracy
Kunal Aggrawal
Available for roles

Experience

Real-world engineering and production systems

AI Engineer

AmlgoLabs — Gurugram
Jan 2026 — Present

Currently leading the development of high-throughput multi-agent systems and automated cloud provisioning pipelines.

⚡
Auto-Apply Engine: Celery/RabbitMQ architecture processing 264K+ jobs in 218 seconds at 4.6 users/sec.
☁️
AWS Provisioning AI: Gemini-driven system extracting NFRs to generate end-to-end Terraform configs and BOMs.
📄
Invoice OCR Pipeline: OCR + LLM parsing ~3,000 Suzuki invoices/day directly into structured schemas.
FastAPILangGraphAWSPostgreSQLCeleryRabbitMQTerraform

Junior AI/ML Intern

AmlgoLabs — Gurugram
Oct 2025 — Dec 2025

Deployed the organization's first on-premise Kubernetes + Kubeflow ML lifecycle system for Concentrix.

📦
MLOps Infra: Model registry, serving, and 4 Prometheus/Grafana monitoring dashboards.
📊
MLflow Tracking: Standardized logging and model comparison across data science teams.
KubeflowKubernetesDockerMLflowPrometheusGrafana

Python Developer Intern

Innovation HacksAI
May 2025 — Sep 2025

Engineered autonomous agent workflows and voice-driven AI platforms to replace manual sales efforts.

🤖
Text-to-SQL Agent: DeepSeek-powered orchestrator achieving ~80% accuracy with auto-retry mechanisms.
📞
Voice AI Pipeline: Vapi + Twilio integration for AI-driven phone outreach, fully replacing manual emails.
FastAPIDeepSeekAutogenVapiTwilioDocker

Selected Work

Deep dive into specific engineering challenges

GitHub ↗

Multi-Agent Customer Support System

Production-grade multi-agent backend using LangGraph + FastAPI. A triage agent routes queries to ReAct-based billing and technical specialist agents that execute tools for ticket history and knowledge-base search.

🔒
Human-in-the-Loop: Approval flow pauses before sensitive actions; supervisors approve/reject via API.
💾
Persistent Memory: PostgreSQL checkpointing and long-term summaries across sessions.
⚡
Async & Tracing: SSE streaming, Celery/RabbitMQ tasks, and full LangSmith trajectory tracing.
LangGraphFastAPIPostgreSQLCeleryRabbitMQLangSmithPython

Skills & Technologies

AI / ML

LangGraph · LangChain · Ollama · Autogen · Scikit-learn · Pandas · NumPy · Neural Networks · CNN

Backend

Python · FastAPI · PostgreSQL · SSMS · REST APIs · JWT · OAuth 2.0 · Celery · RabbitMQ · Docker

MLOps

Kubeflow · Kubernetes · MLflow · Prometheus · Grafana · GitHub Actions

Cloud & Infra

AWS (EC2, ALB, RDS, S3, VPC, ECR, Route 53, CloudWatch, IAM) · Nginx · Fail2Ban · SSH hardening · Terraform

Education

Master of Computer Applications (MCA)

University of Delhi, Dept. of Computer Science — New Delhi
Sep 2023 – Aug 2025 CGPA 8.45

Bachelor of Science (B.Sc.)

Sri Dev Suman University — Tehri Garhwal, Uttarakhand
Jul 2020 – Aug 2023 CGPA 7.87

Contact

Open to AI/ML engineering and backend roles. Let's talk.

Email aggrawalk970@gmail.com
LinkedIn kunal-aggrawal ↗
GitHub aggrawalk970-cyber ↗
Phone +91-7417738058