filmIQ builds the intelligence layer for physical operations, starting with film and television production.
filmIQ is building the labeling apparatus that lets the industry measure what its standard outcome metrics are systematically corrupted to conceal. The thesis transfers from production to every other complex physical operation where knowledge work holds the system together. This page lays out the thesis, the proof, and the honest limits of what we've built so far.
The thesis.
Complex physical operations are coordinated by a layer of knowledge work that connects information systems to what's actually happening in the world. Line producers reading daily production reports. Unit production managers cross-referencing call sheets and cost reports. Construction project managers triangulating schedules against field reports. Clinical trial coordinators reconciling enrollment data against protocol.
This category of knowledge work has three defining features. It is performed by experienced practitioners whose judgment depends on patterns they've accumulated across years of operational reality. It is entangled with a physical world that does not wait for analysis. And it has, until recently, been invisible to software because the inputs are unstructured documents and the outputs are human decisions.
AI capability has now reached the threshold where this category of work can be meaningfully augmented. The enterprise AI wave of the last three years solved the easier version of the problem: knowledge work where the inputs and outputs are both digital, where the feedback loop is tight, and where the domain can be understood from the artifact alone. The harder version is knowledge work entangled with physical operations, where the inputs are messy, the feedback loops are long, and the domain context lives in the heads of practitioners rather than in any system.
The harder version is where the largest economic opportunities sit. It is also where capability alone is not enough. Intelligence without operational context is unreliable in high-stakes environments. Winning these categories requires something capability alone can't provide: structured operational intelligence built through expert elicitation, validated against real operational data, and embedded in workflows that don't ask the practitioner to change what they're already doing.
filmIQ is building that intelligence layer, starting with film and television production.
The buyer is the capital allocator carrying exposure to a complex execution they cannot directly observe, currently priced off corrupted outcome data because nothing better exists. Studios, networks, streamers, financiers, and completion bond companies share this position. They underwrite projects whose execution quality they cannot directly read from the documents the projects produce. filmIQ is built for them.
filmIQ does not sell access to operational data. It sells the contextual framework that makes operational data interpretable, predictive, and trustworthy in a domain where outcome data alone cannot get you there. The framework is what does not exist anywhere else in the industry, and it is the moat.
Why film and television production is the right beachhead.
Production is the category where operational drift, expert judgment, and document-based coordination all concentrate in a time-bounded environment that produces rich data and measurable outcomes.
High operational complexity, short time horizon.
A single scripted series production runs 25 to 100 shooting days. Every production is a complete operational cycle with a defined beginning, middle, and end. Pattern learning happens on a timescale of months, not years.
Document-rich coordination.
Call sheets, DPRs, cost reports, hot costs, schedules, scripts, time cards, script supervisor dailies. The operational reality of every production is already being written down. The data exists. It simply isn't being read at the scale or pace AI can read it.
Outcomes that matter and can be measured.
Productions have concrete outputs: footage shot, pages completed, budgets spent, schedules held, creative scope delivered. The variables that matter are legible, even when the compounding drift between them is not.
Deep tacit knowledge that has never been structured.
Line producers and UPMs carry pattern libraries in their heads that no system has ever captured. The opportunity to elicit, structure, and operationalize that knowledge is a once-per-category event. Whoever does it first wins institutional memory for the industry.
These four properties are not unique to production. They are shared by construction, clinical trials, logistics coordination, energy infrastructure, and every other complex physical operation where knowledge work holds the system together. Production is the beachhead. The thesis transfers.
Why the moat compounds.
Four assets compound as filmIQ deploys.
The architectural inversion.
Standard production software (legacy desktop tools and current AI-era platforms alike) frames its product around "single source of truth." That framing is structurally wrong for what filmIQ does. Production data is not one underlying reality recorded by multiple imperfect systems. It is multiple legitimate truths held simultaneously by parties with different responsibilities and different time horizons. filmIQ holds those layers separate, reads the gaps between them, and detects what moves through them as the signal. Every SSOT claim in the competitive landscape is scoped to a single layer. Nobody claims SSOT across operations, creative vision, and contractual constraints because nobody can without the layered-truth architecture. The architectural inversion is what unlocks the institutional buyer category that operations software was never designed to serve.
The pattern library.
Every production filmIQ deploys on contributes to the library of patterns telemetry can detect. Every time our research panel validates a new pattern, the library gains reach. Every time a practitioner flags a false positive or false negative, the library gains calibration. Capability alone does not produce this library. Neither does raw data. It is the pairing of structured practitioner knowledge with operational data from real productions, and it gets harder to replicate with every deployment.
The research panel.
Our panel members are working producers with more than a hundred combined production credits. They are the research instrument for pattern library construction, not an advisory board in the conventional sense. Their judgments at temporal inflection points during productions are the labeling mechanism that produces the training signal outcome data cannot provide. Their access, credibility, and willingness to engage is a function of relationships built over years, not something a competitor can assemble in a round of fundraising.
Institutional memory.
Every production we deploy on generates data that, with appropriate permissions, becomes part of the structured corpus filmIQ can reason over. Over time, the accumulated record of how productions actually behave becomes an asset the industry does not currently have and does not know how to build. The downstream opportunity includes studio archive monetization, portfolio-level intelligence for completion bond companies, and a reference corpus for the industry at large.
What we've built.
filmIQ has one foundational pilot and one active deployment. What follows is the honest state of evidence.
First live deployment.
An end-to-end deployment on a scripted series. The platform ingested several hundred documents in heterogeneous formats with high first-pass parsing success. Operational drift registered in the schedule layer within the first week of production. Financial reporting stayed clean across the run, an early observation of the absorption pattern that subsequent architecture is built around. Specific performance numbers are reserved for due-diligence conversations rather than public claims.
The finding is the middle number. The financial metrics stayed clean because the production absorbed the drift through reduced creative scope. The roughly 12-day lead time between the first operational signal and the point at which the drift would have become visible in financial reporting is the value of the product. It is the interval during which creative compression is still reversible. The figure itself is pending engineering verification. Building the broader evidence base across productions is part of what subsequent deployments contribute to.
Active deployment.
telemetry is currently live on an active feature production. This deployment is our first against a feature rather than a series, and the first where telemetry is running in real time alongside active production decisions. We'll report on outcomes publicly when the production wraps and our producing partners have reviewed what's appropriate to share.
What we learned going from V1 to V2.
The initial pilot validated the detection thesis. It also exposed clear product limits. The V1 intelligence layer delivered findings through a dashboard that required practitioners to remember to open it. The findings themselves, at times, contained speculation and pleasantries that destroyed credibility with experienced producers. Both problems were foundational, not superficial.
V2 addresses them directly. The findings now arrive through email, not a dashboard the practitioner has to seek out. The language is terse, evidence-backed, and free of speculative content. Credibility with practitioners is the binding constraint of this category, and the V1-to-V2 transition is what we did to meet it.
What we are not claiming.
We have tried to be careful about the edges of our evidence.
These are the edges. Within them, we are confident. Beyond them, we will be honest as the evidence comes in.
Where this goes.
The seed round is the infrastructure phase. It funds the pattern library build-out, the research panel engagement, and the next set of deployments that turn the initial pilot into a validated track record.
Series A is the expansion phase. It funds the first studio-scale deployments, the portfolio-level intelligence product for completion bond companies, and the beginning of the adjacent-category work that proves the thesis transfers beyond production.
The categories immediately adjacent to production are construction, clinical trials, and logistics. The technical architecture and the elicitation methodology port directly. The decision on which adjacency to enter first will depend on deployment signal, partner access, and where the pattern library is strongest at that moment.
We are not pitching the adjacencies to seed investors. We are pitching the beachhead. The adjacencies are the reason the beachhead is worth winning.
The round.
filmIQ is raising its seed round to fund the pattern library, the next cohort of deployments, and the commercial infrastructure for the V2 product.
Additional detail, including round size, terms, lead status, and current commitments, is shared directly with qualified investors. We prefer to do this in a short conversation rather than on a public page.