Montaser Hussam
AI/ML engineer based in Gaza, Palestine, working remotely with teams across time zones.
Working method
Scope narrow first. A voice or agent feature is broken into the smallest slice that can be demoed end to end, then widened once that slice is measured and working — not designed fully up front and built in one long push.
Every architectural choice states its cost. A faster integration that cannot survive a duplicate webhook, or a lower-latency endpointing threshold that misfires in a noisy room, gets named as a trade-off rather than presented as free.
What he is looking for
Teams building a real-time voice product, an agent-based workflow, or a production LLM feature that needs to survive contact with real users and real network conditions — not a proof of concept that stays a proof of concept.
Experience
Muhawir — Founder & engineer
Designed and built a bilingual real-time voice interview coach end to end, from the WebRTC transport layer to post-call scoring and billing.
Independent AI/ML engineer
Delivered agent pipelines and LLM-backed features for early-stage products, focused on shipping working systems on realistic timelines.
BSc, with distinction
Undergraduate degree with a focus on software engineering and applied machine learning.