Marketplace Explore 6,000+ AI-powered automations to scale your business.

AI-Powered Video Research Platform

Deeptech video – Autonomous research & video pipelines

Agentic R&D platform automating experimentation, evaluation and video-to-video generation pipelines.

Research steps orchestrated

296

Historical tests integrated

Autonomous research loop

From literature monitoring to evaluated video pipelines, in one orchestrated cycle

Project Details

Client

Inevitable

Sector

Deeptech video – Autonomous research & video pipelines

Year

2026

Key Technologies

LangGraphXGBoostComfyUIFastAPI

Mission

Lead AI Product Engineer - Autonomous Research & Video Pipelines

Autonomous research & video pipelines

Full remote, France. English.

Design and development of an autonomous research platform for video-to-video transformation (3D/game engine rendering → photorealistic video). Orchestration of an iterative 21-step research cycle via LangGraph, from literature monitoring to video pipeline execution, with automatic evaluation and continuous learning. Integration of an ML recommendation engine (XGBoost + SHAP) to suggest optimal configurations and produce explainable policies. Full-stack development: React 18 front-end (workflows, dashboards, evaluation) and FastAPI back-end (orchestration, REST API, multi-backend integration).

Contexte & Environnement

Internal R&D platform dedicated to automating the full cycle of monitoring, experimentation and capture of learnings in video-to-video. Autonomous research loop: hypothesis → Scholar scraping → method extraction → planning → execution → evaluation → learning → refinement. Integrated existing pipeline: wrapping of a ComfyUI pipeline driven by Claude agents (Builder Sonnet + Reviewer Opus), with 296 historical tests and persistent learnings.

Team

Lead AI Product Engineer – full-stack, ML and agents

Environment

Full remote, France – React and FastAPI, LangGraph agents, ComfyUI video pipeline on Google Cloud

Objectifs Cles

1Automate the research cycle via 21-step LangGraph orchestration with optional human gates.

2Deliver ML-driven recommendations: candidate generation, XGBoost scoring, top-K ranking and SHAP explanations.

3Capitalize on knowledge: structured extraction from Google Scholar, knowledge records, derived policies (hard bans, soft preferences, SHAP-derived).

4Evaluate on multiple criteria: automatic metrics (PSNR, SSIM, temporal stability, flicker) and subjective human evaluation.

Technologies & Infrastructure

React 18TypeScriptFastAPISQLAlchemyCeleryLangGraphLiteLLMClaudeXGBoostSHAPOptunaComfyUIOpenCVFFmpegPostgreSQLBigQueryRedisDockerGoogle Cloud

Ready to Achieve Similar Results?

Let's discuss how we can drive measurable impact for your organization.

Other Business Cases

Explore how we drive results across industries

AI Skills Assessment at Scale

Banking

Skills ontology, large-scale assessments, ML scoring and personalized learning pathways.

Learn more

AI Architecture for Deliverable Automation

Rail Industry

Data and AI architecture to automate deliverable production, quality control and traceability.

Learn more

AI Technology Intelligence & Literature Review

Insurance

Full-stack technology intelligence platform with source analysis, literature review and AI benchmarking.

Learn more