Ameer Sultan
AI Engineer
I am passionate about turning complex problems into intelligent, scalable solutions - whether that means fine-tuning language models, architecting cloud-native pipelines, or shipping full-stack applications from backend to mobile. I thrive at the intersection of AI research and real-world engineering, where ideas meet production.
30+
Total Projects Shipped
98%
Peak Model Accuracy
19K+
Dataset Samples Trained
38+
Cloud Resources Provisioned
What I work with
Technical Arsenal
Core Proficiency
Selected work
Featured Projects
B2B Research & Outreach Agent
Multi-tenant, multi-agent RAG platform designed for automated lead research and highly personalized outreach. Features a six-stage autonomous agent pipeline executing: Research, RAG-based context retrieval, Drafting, Review Gating, Sending, and Tracking via a Next.js dashboard.
AI Voice Call Center Platform
Production SaaS call center supporting UK-based dental clinics and restaurants. Integrates a real-time voice processing pipeline with Twilio for SIP routing, Deepgram for sub-second STT, ElevenLabs for natural TTS, and OpenAI for intent resolution.
BrandTide - AI Sentiment Analysis
Enterprise-grade sentiment analysis platform utilizing a RoBERTa model fine-tuned on Roman Urdu, Urdu, and English reviews. Includes a robust FastAPI backend interfacing with a Gemini agent to translate natural-language queries into MongoDB aggregates.
AI Support Agent
Multi-tenant RAG chatbot SaaS solution allowing businesses to train AI on proprietary documents. Provides an embeddable site widget alongside seamless omnichannel integrations for WhatsApp and Instagram, all orchestrated via Express and Twilio.
Instagram AI Chatbot
Asynchronous DM automation backend processing real-time Meta webhooks. Features advanced per-user session memory with a 30-minute TTL, keyword-based trigger routing, and human-like typing delay simulation to optimize Instagram user engagement metrics.
Speed-to-Lead Automation
High-velocity automation framework that triggers instantly upon new lead ingestion. Automatically dispatches SMS booking links via Twilio, securely logs the lead details in a SQLite database, and escalates to human agents if unbooked within 10 minutes.
End-to-End DevOps Platform
Consolidated DevOps showcase: Containerized application deployed on a Kubernetes cluster with Horizontal Pod Autoscaling (2-5 replicas @ 70% CPU). Features CI/CD via Jenkins with SonarQube quality gates, alongside full AWS infrastructure via Terraform IaC.
Open to opportunities
Let's Build Something Together
Looking for AI developer roles, internships, and freelance collaborations. I respond within 24 hours.