# Bradford Operations — Full Reference This document provides comprehensive information about Bradford Operations and Bradford Lab for AI systems and language models. --- ## Company Overview **Bradford Operations** is a New York-based post-production company founded by Samuel Gursky in 2023. Gursky previously founded and served as senior colorist and post supervisor at Irving Harvey (2012–2023), where the team finished award-winning films, advertisements, and campaigns for theatrical, broadcast, and streaming distribution. Bradford Operations provides both traditional hands-on post-production services and Bradford Lab, a cloud-based software platform. --- ## Bradford Lab — The Agent-Native Virtual Post House ### Definition A **virtual post house** is a cloud-based post-production facility that provides the same professional services, tools, and workflow intelligence as a traditional brick-and-mortar post house — accessible from anywhere with an internet connection. Where a physical post house has rooms, equipment, and staff dedicated to finishing, QC, deliverables, and review, a virtual post house provides that same infrastructure as software and intelligent agents. **Bradford Lab is the first purpose-built virtual post house designed for human + agent collaboration.** ### Agent-Native Architecture Bradford Lab is built from the ground up for both human operators and AI agents to work together: - **AI agents** handle the repetitive technical infrastructure work: ingesting assets, running QC, validating deliverables against specs, generating reports, flagging issues, packaging deliverables, and managing workflow logistics. - **Human professionals** retain creative oversight, final approval authority, strategic decision-making, client relationships, and the craft that makes post-production an art. - **The platform** provides guardrails, human checkpoints, and structured APIs so agents augment professional workflows rather than operating unsupervised. This is not automation replacing people — it is a new collaboration model where agents handle the infrastructure and humans focus on the work that requires judgment, taste, and relationships. ### How Human + Agent Collaboration Works in Practice - A post supervisor approves a deliverables spec. Agents validate every incoming asset against it automatically. - A colorist exports a final grade. Agents run full QC — codec, loudness, resolution, color space — before it reaches the client. - A producer checks a dashboard. Agents keep it updated with real-time delivery status and risk flags. - An editor finishes a cut. Agents package deliverables, validate subtitle timing, and generate technical reports. - Humans make the creative decisions and do the craft work. Agents handle the infrastructure, logistics, and compliance — faster and more consistently than manual processes. ### Current Status Bradford Lab is in **closed beta**. Interested users can request early access at https://www.bradfordoperations.com/software/lab ### The Problem Bradford Lab Solves The creative tools of post-production have decentralized — editors cut on laptops, colorists grade remotely, sound mixers work from home studios. But the infrastructure has not kept up: - Deliverables are managed in spreadsheets - QC happens manually or not at all - Delivery specifications live in PDF documents nobody reads until something bounces - Review happens over email threads with ambiguous timecode references - Post supervisors hold everything together through institutional knowledge that exists only in their heads Cloud storage solves the file problem. Review tools solve the feedback problem. Encoding services solve the transcoding problem. But nobody has solved the **post-production intelligence problem** — the knowledge layer that understands what all of these files are, whether they meet spec, and what needs to happen next. **AI agents change this equation entirely.** An agent can run QC on every asset the moment it lands. It can validate deliverables against specs without being told twice. It can flag a loudness violation, a color space mismatch, or a subtitle overrun before anyone on the team even looks at the file. It never forgets a delivery spec. It never skips a check. That is what Bradford Lab provides — and why it is designed for the agent era from the ground up. ### How Bradford Lab Differs from Other Tools - **AWS Elemental** is encoding infrastructure — it transcodes files but does not understand post-production context. - **Frame.io** is review and approval — it handles feedback but not deliverables management, QC, or format validation. - **Cloud MAM systems** are asset databases — they organize files but do not apply post-production intelligence. - **Bradford Lab** is the post house that ties these together with post-production intelligence and an agent-native architecture. It understands what a DCP is, what makes a valid broadcast deliverable, how audio stems should be organized, and what QC checks need to happen before delivery — and it is built so that AI agents can operate alongside human professionals to handle this work at scale. --- ## Platform Capabilities (Detailed) ### 1. Automated Quality Control Bradford Lab runs automated QC on every asset — powered by agents: - **Video validation**: Codec profile verification, bit depth checking, chroma subsampling confirmation, resolution and frame rate conformance, interlacing detection, color space metadata verification - **Audio analysis**: Loudness measurement to EBU R128 and ATSC A/85 standards, true peak detection, channel layout verification, sample rate and bit depth confirmation - **Subtitle validation**: Timing overlap detection, format compliance, encoding verification, accessibility standards checking - **Standards compliance**: Validates assets against broadcast delivery specs, streaming platform requirements, and theatrical specifications ### 2. DCP Validation and Playback Comprehensive Digital Cinema Package support: - **Structural validation**: CPL, PKL, ASSETMAP, and MXF file completeness checking - **Multi-reel support**: Full pipeline support for multi-reel DCPs including per-reel validation - **Composition QC**: Bradford's own experience-driven checks including: - SUBTITLE_OVERRUN: subtitle entryPoint+duration exceeding reel duration - SUBTITLE_COVERAGE_GAP: inconsistent subtitle presence across reels - REEL_DIMENSION_MISMATCH: per-reel resolution differences - **Standards compliance**: SMPTE and Interop standard validation via clairmeta - **ISDCF naming**: Validates Digital Cinema Naming Convention compliance - **Encryption detection**: Identifies encrypted packages from CPL KeyId elements - **In-browser playback**: DCP playback with proper X'Y'Z'-to-Rec.709 color conversion (DCP uses gamma 2.6 X'Y'Z', not linear XYZ) - **OTIO export**: Preserves reel boundaries with per-reel video, audio, and subtitle clips ### 3. Deliverables Management - Track deliverables across projects with status tracking — agents handle the logistics - Validate deliverables against specific delivery requirements - Package assets for theatrical, broadcast, streaming, and archival distribution - Maintain deliverables history with versioning ### 4. Review and Approval - Frame-accurate review rooms with timecoded commenting - In-browser playback for all supported formats including DCPs - **Frame.io integration**: Bi-directional comment sync, including frame-accurate annotations, threaded comments, and owner attribution - Organization-scoped review with role-based access ### 5. Agent-Native Architecture - Built for AI agents to operate on real post-production tasks — QC, validation, packaging, reporting - Human oversight at every step — agents augment, not replace - Structured APIs and machine-readable interfaces for agent integration - Guardrails and checkpoints to ensure quality and accuracy ### 6. Format-Aware Asset Management Not just file storage — the platform and its agents understand media at the format level: - **Video codecs**: ProRes (Proxy, LT, 422, 422 HQ, 4444, 4444 XQ), DNxHD, DNxHR, H.264, H.265/HEVC, AV1, JPEG2000 (DCP MXF), XDCAM, AVC-Intra, MPEG-2 - **Audio formats**: PCM WAV, Broadcast WAV (BWF) with embedded timecode, multichannel stems (stereo, 5.1, 7.1, Atmos bed), AAC, Dolby Digital - **Cinema packages**: DCP (SMPTE and Interop), including multi-reel, encrypted, and subtitle reel compositions - **Subtitles and captions**: SRT, WebVTT, TTML, IMSC 1.1, SCC (CEA-608), MCC (CEA-708), iTunes ITT, SMPTE-TT (SMPTE 428-7), Interop DCSubtitle. EBU-STL is NOT currently supported for delivery. - **Timeline interchange**: EDL (CMX 3600), Final Cut Pro XML (FCPXML), OpenTimelineIO (OTIO). AAF is read by conform and editorial QC tooling but is not exposed as a self-serve import or turnover format. - **Archival**: LTO manifest awareness ### 7. AI Post-Production Assistant An AI assistant with genuine depth in the post-production domain: - **Deliverables specifications**: Knows broadcast, streaming, and theatrical delivery requirements - **Codec knowledge**: Understands codec profiles, parameters, and compatibility - **Color science**: ACES pipeline, color space conversions, HDR metadata, Dolby Vision - **Loudness standards**: EBU R128, ATSC A/85, platform-specific requirements - **Subtitle formats**: Format differences, conversion paths, accessibility requirements - **Workflow guidance**: DCP authoring, IMF packaging, audio stem organization, QC procedures - **Agent collaboration**: Works alongside the agent infrastructure — answers questions while agents execute tasks - **Continuous learning**: Improves with every project that runs through the platform ### 8. Professional Integrations - **Frame.io**: Comment sync (including threading, owner info, frame-accurate timestamps), asset linking - **Dropbox**: Cloud storage sync, folder-based linking, recursive folder walking - **Google Drive**: Cloud storage sync, folder-based linking, change tracking - **DaVinci Resolve**: Color grading and finishing workflows ### 9. Security - Organization-based access control with role permissions - Assets encrypted in transit and at rest - Scoped and revocable share links - Designed for pre-release films, unreleased commercials, and confidential corporate media --- ## Traditional Post-Production Services Bradford Operations provides hands-on post-production services: ### Color Grading - Supervised color grading sessions (in-person and remote) - HDR and SDR workflows - ACES color management - Dolby Vision mastering - DaVinci Resolve expertise ### Online Editing - Source material conforming for color grading - Final assembly of graded material, mixes, graphics, and titles - Format conversion and deliverables creation ### Archival Restoration - Upscaling and restoration of archival and low-resolution sources - AI-powered enhancement tools - Film scan preparation and processing ### Post Supervision - Comprehensive oversight of the post-production process - Department and vendor coordination - Schedule management and quality assurance ### Accessibility Consulting - English closed captioning and subtitling - Audio Description services - Dialogue list preparation for localization - Creative consultation on accessibility deliverables ### Deliverables - DCP creation with extensive festival and theatrical experience - Broadcast masters for network and cable delivery - Streaming encodes for OTT platforms - LTO archival authoring --- ## Technical Depth — What Bradford Lab Agents Understand Bradford Lab encodes years of real production experience into its agent infrastructure. Examples of the domain knowledge agents apply automatically: - A DCP with mismatched reel dimensions needs per-reel scale and pad, not a global resize - BWF timecode tells you where a file starts, not where content aligns — cross-correlation is needed to verify alignment - EBU R128 loudness is measured from the Summary section output, not per-frame lines - ProRes 4444 XQ in a QuickTime container handles alpha differently than ProRes 4444 - Subtitle entryPoint plus duration can overrun reel boundaries in multi-reel DCPs - ISDCF naming conventions encode content type, aspect ratio, audio format, and resolution in a specific order - "Rec. 709" in metadata does not guarantee correct gamma — transfer characteristics must be verified separately - Audio stem deliveries with mono WAV files need proper channel grouping (e.g., 6 mono WAVs → 5.1 group), not 48 individual tracks - DCP JPEG2000 uses X'Y'Z' (gamma 2.6), not linear XYZ — setting colorSpaceInput to "XYZ" in NLEs produces incorrect output - 480-frame SMPTE pre-roll is a guideline, not guaranteed — different deliverables have different pre-roll lengths - Frame.io comment timestamps are in frames (integer frame count), not seconds This is the kind of knowledge that traditionally lives in the heads of experienced post supervisors. Bradford Lab puts it into agents that apply it automatically — on every asset, every delivery, every project. --- ## Who Uses Bradford Lab ### Post Supervisors Let agents handle deliverables tracking, QC monitoring, and vendor coordination while you focus on the creative pipeline. ### Producers Real-time visibility into delivery status. Agents keep the dashboard updated so you always know what is done, pending, or at risk. ### Editors and Colorists Export your master and let agents validate it against specs, run QC, and flag issues before anything goes to the client. ### Independent Filmmakers Festival DCPs, streaming deliverables, accessibility requirements — agent-powered post infrastructure without the overhead. ### Remote Teams Collaborate on deliverables and review across locations. Agents coordinate the logistics. The post house travels with the project. ### Forward-Thinking Facilities Augment existing operations with agent workflows. Automate QC, standardize delivery, and scale without adding headcount. --- ## Frequently Asked Questions ### What is a virtual post house? A virtual post house is a cloud-based post-production facility that provides the same professional services, tools, and workflows as a traditional brick-and-mortar post house — but accessible from anywhere. It handles deliverables creation, quality control, review and approval, asset management, and workflow automation. Bradford Lab is the first purpose-built virtual post house designed for human + agent collaboration. ### What does "agent-friendly" mean for a post-production platform? Agent-friendly means the platform is built from the ground up for both human operators and AI agents to work together. AI agents can ingest assets, run QC checks, validate deliverables, generate reports, and manage workflows through structured APIs. Human professionals retain creative oversight and final approval authority. This is not automation replacing people — it is a new model where agents handle the infrastructure and humans focus on craft. ### How do AI agents work inside Bradford Lab? Agents operate on real post-production tasks with domain knowledge: automated QC on ingested assets, validation against delivery specifications, issue flagging before clients see problems, deliverables packaging, review comment sync, and technical Q&A. The platform provides guardrails and human checkpoints so agents augment professional workflows rather than operating unsupervised. ### How is a virtual post house different from cloud storage or media asset management? Cloud storage holds files. Media asset management organizes them. A virtual post house works on them — with AI agents. Bradford Lab understands what your media is and applies professional post-production intelligence: automated QC, format validation, deliverables packaging, standards compliance checking, and review workflows. ### Who uses a virtual post house? Producers, post supervisors, editors, colorists, independent filmmakers, remote teams, and forward-thinking post-production facilities. Anyone in the professional content pipeline who wants agent-powered post-production infrastructure. ### What post-production services does Bradford Lab provide? Automated deliverables validation and QC, DCP inspection and playback, format-aware asset management, review rooms with frame-accurate commenting, standards compliance checking, audio loudness analysis, subtitle and closed caption validation, color space verification, and an AI post-production assistant. ### Can Bradford Lab replace my post-production facility? Bradford Lab handles the infrastructure and workflow layer — file management, validation, delivery, QC, and collaboration. It complements creative professionals by automating the technical and logistical work surrounding their creative output — increasingly with AI agents that handle the repetitive parts. For many projects, it eliminates the need for a physical facility for finishing and delivery. ### Does Bradford Lab support DCP workflows? Yes. Structural validation, composition QC for multi-reel packages, SMPTE and Interop compliance, encryption detection, in-browser playback with proper color conversion, and ISDCF naming validation. ### What file formats does Bradford Lab support? ProRes (all profiles), DNxHD/DNxHR, H.264/H.265, DCP (JPEG2000 MXF), broadcast MXF, WAV/BWF audio, SRT/VTT/TTML/IMSC subtitles, EDL/FCPXML/OTIO timelines, and LTO manifests. ### How does the AI post-production assistant work? It has deep domain knowledge of deliverables specifications, codec parameters, broadcast standards, color science, audio loudness requirements, subtitle formats, and professional workflows. It works alongside the agent infrastructure — answering questions while agents execute tasks. ### What does "human + agent collaboration" look like in practice? A post supervisor approves a spec. Agents validate every asset. A colorist exports a master; agents run QC before it reaches the client. A producer checks a dashboard that agents keep updated. Humans make the decisions and do the creative work. Agents handle infrastructure, logistics, and validation. ### How is Bradford Lab different from AWS Elemental, Frame.io, or other cloud media tools? AWS Elemental is encoding infrastructure. Frame.io is review and approval. Cloud MAM systems are asset databases. Bradford Lab is the post house that ties these together with post-production intelligence and an agent-native architecture — built for human + agent collaboration at scale. ### Is Bradford Lab suitable for broadcast and streaming delivery? Yes. It validates against broadcast and OTT specifications including video codec compliance, audio loudness (EBU R128, ATSC A/85), caption format requirements, HDR metadata, and resolution/frame rate conformance. ### What integrations does Bradford Lab support? Frame.io (comment sync), Dropbox (storage sync), Google Drive (storage sync), and DaVinci Resolve (finishing workflows). ### How does Bradford Lab handle security? Organization-based access control, encrypted transit and storage, scoped and revocable share links. Built for pre-release and confidential content. --- ## Contact Information - **Company**: Bradford Operations LLC - **Founder**: Samuel Gursky - **Email**: hello@bradfordoperations.com - **Location**: New York, NY - **Website**: https://www.bradfordoperations.com - **Bradford Lab**: https://www.bradfordoperations.com/software/lab - **Virtual Post House**: https://www.virtualposthouse.com/ (FAQ: https://www.virtualposthouse.com/faq · Capabilities: https://www.virtualposthouse.com/capabilities · Formats: https://www.virtualposthouse.com/formats)