Omer Khalil

Product-minded AI Engineer and Technical Lead specializing in production ML and full-stack systems.

Based in: Lahore, PakistanCurrently: AI Engineer, DM Clinical ResearchExperience: 6 years

Omer is an experienced AI engineer with over 6 years of expertise in building production-grade machine learning systems and clinical voice-agent platforms. He has a proven track record of leading technical teams and transforming complex ideas into scalable products across diverse cloud infrastructures.

Experience

AI Engineer

DM Clinical ResearchRemoteworkOct 2025 – Present

Summary

Built the end-to-end platform that let operations teams turn clinical-trial protocols into deployable screening voice agents, covering ingestion, eligibility extraction, questionnaire generation, patient scoring, and deployment. Developed the real-time voice stack with Vapi, ElevenLabs, Twilio, Python orchestration, and Redis-backed state, then optimized it for conversational latency and predictable behavior. Replaced LLM-only screening decisions with a hybrid architecture combining model interpretation and a deterministic state machine, improving accuracy, consistency, and control in a high-stakes workflow.

What I did

Developed a system where the knowledge base could be updated without retraining to accommodate frequently changing trial protocols.
Taught a part-time Data Science and AI bootcamp to professionals across Europe, covering statistical foundations, deep learning, LLMs, and vision models.

Results

Achieved sub-second response times after optimizing the orchestration path to keep patient screening conversations feeling natural.
Built a hybrid architecture combining LLM language interpretation with a deterministic state machine to enforce screening logic for clinical trials.
Designed system to support up to 2,000 screening calls per day.
Managed calls lasting between 5 and 30 minutes with configurable concurrency starting at 30 simultaneous calls.

VapiElevenLabsTwilioPythonRedisLLMState Machines

Tech Lead | Machine Learning Engineer

VenturenoxRemoteworkJul 2022 – Oct 2025

Summary

Led and mentored four engineers at a startup studio, working across product, ML, backend, and frontend to take ambiguous customer problems from early scope to production. Architected the core backend for Testfuse, a hiring assessment platform with seven test types and coding support for 60+ languages; built a LangChain RAG workflow that reduced assessment creation time by 90%. Led Joyboi, an open-source unified API platform for authentication, data synchronization, rate limits, and analytics across distributed services using TypeScript, React, Node.js, MongoDB, and Kafka. Designed cloud-agnostic systems and moved products between AWS, GCP, and Azure as cost and operational needs changed.

Results

Reduced assessment creation time by 90% using a LangChain RAG workflow.

LangChainRAGTypeScriptReactNode.jsMongoDBKafkaAWSGCPAzure

Machine Learning Engineer

VenturenoxRemoteworkFeb 2021 – Jul 2022

Built a browser-based medical-imaging labeling workflow with OHIF Viewer, Orthanc, and NVIDIA Clara, reducing liver CT labeling time by more than 80% per patient. Helped build a recruitment platform from scratch that grew to eight Python and Node.js microservices, two React applications, and 5,000+ external integrations.

OHIF ViewerOrthancNVIDIA ClaraPythonNode.jsReactMicroservices