Realtime AI Voice Agent
Designed and deployed an intelligent, low-latency voice agent capable of natural two-way conversation in real time using various technologies.
Bilal S. is an AI Engineer with several years of experience in building high-performance, production-grade systems across full-stack development, cloud platforms, and applied AI. His core stack includes .NET (ASP.NET MVC/Web API), Angular, React, and Node.js, used to deliver scalable applications with clean architecture. He integrates PostgreSQL, SQL Server, and Redis, employing event-driven patterns like CQRS and MediatR. Bilal also utilizes OpenAI, Whisper, and AWS Bedrock for AI solutions, focusing on real-time interaction through WebRTC and Twilio, and ensures robust deployment with Docker and monitoring via OpenTelemetry and Grafana. At Greybeard – Halifax Fans, Bilal enhanced authentication systems for over 5,000 users and improved selector performance by up to 900%. He also automated CAD workflows and reduced database calls by 20%. At Oaks Street Technologies, he resolved 17 critical bugs in a week and implemented real-time capabilities using SignalR, strengthening security with JWT/role-based authorization. Bilal holds a BS in Computer Science from FAST NUCES. He is well-suited for roles that involve designing, deploying, and scaling intelligent software systems, particularly those requiring advanced AI integration and cloud-based solutions.
Designed and deployed an intelligent, low-latency voice agent capable of natural two-way conversation in real time using various technologies.
Built an ingestion pipeline that detects & extracts text/tables from PDFs with Textract, chunks/embeds content, and indexes to OpenSearch for retrieval.
Deployed a scalable RAG backend using Bedrock Knowledge Bases over Confluence/SharePoint dumps.
Loads a public URL, cleans & chunks content, embeds via Gemini, stores in Pinecone, and answers questions with source-grounded responses.
Interactive 3D avatar with GLB/GLTF clothing overlay, draggable camera, animations, and responsive UI.
Implemented conditional GANs for age progression/regression with identity preservation.
Persona-aware email drafts with tone/length controls, dynamic snippets from a prospect’s site, and deliverability checks.
Simulated beacon chain epochs, validator rewards/penalties and slashing rules; lightweight dashboard to visualize head/finality and staking performance on testnets.
LSTM next-word prediction trained on Shakespeare; Streamlit UI and a small Flask/uvicorn API for programmatic access.
Other vetted developers with similar skills and experience

AI Engineer
0-2 years experience
14 days risk-free trial · No commitments · We handle contracts and payroll