Audio reduction noice

I have a single phone-recorder file captured in a crowded public space. The speakers I need are buried under loud background noise and I must be able to present the cleaned result. You will receive the original recording in its native format. I need you to: • Remove or attenuate traffic hum, chatter, and other broadband noise without introducing artefacts • Isolate the primary voices so every word can be clearly understood at normal listening volume • Deliver the restored track in WAV and a lightly compressed copy in MP3, along with a short log of the plug-ins, filters, or AI tools you applied and the settings that materially affect authenticity Acceptance criteria – Dialogue is comprehensible on standard headphones with no need to boost volume beyond 60 % – No noticeable digital distortion, phasing, or pumping introduced by the clean-up – Your process notes are detailed enough that an independent engineer could replicate the key steps Preferred tools include iZotope RX, Adobe Audition, or similar forensic-grade suites, but feel free to use any workflow that meets the above standards. Turnaround is flexible within a few days, yet let me know if you can deliver sooner.