02 /Research & Blog
Research and notes
Papers, benchmark results and engineering notes from the team.
Featured
Benchmarks · Oct 6, 2026 VideoDB leads a 100-hour action retrieval benchmark against AWS Nova and Twelve Labs
A head-to-head retrieval benchmark, and what it means for robotics teams choosing what their next policy should learn from.
Essay · Oct 27, 2026 How a robot remembers
“Find my bottle and bring it to me.” A robot can only do this if it remembers what it has seen. This is a plain explanation of memory for robots and agents: the frames it keeps, the indexes that reach them, and how every task leaves the memory better than it found it.
Papers · Aug 8, 2026 Search over the Visual World
Persistent visual memory, layered indexes and source-grounded evidence for agents over cameras, screens and archives.
Essay Oct 27, 2026 How a robot remembers Ashutosh Trivedi Engineering Oct 26, 2026 How to evaluate a video AI platform: a PoC checklist Aditi Priya Engineering Oct 26, 2026 Understanding robotics datasets: inside LeRobot, RLDS and MCAP Om Gate Engineering Oct 22, 2026 Frame access that keeps research moving Samuel Alexander Essay Oct 14, 2026 What is Physical AI? AI that acts in the real world Aditi Priya Benchmarks Oct 6, 2026 VideoDB leads a 100-hour action retrieval benchmark against AWS Nova and Twelve Labs Sankalp Nagaonkar, Samuel Alexander Agents Oct 5, 2026 Agentic video analysis: what works and what doesn't Aditi Priya Agents Oct 5, 2026 Why a search tool is not enough for an LLM working with video Aditi Priya Essay Sep 18, 2026 A query engine for robot video Ashutosh Trivedi Benchmarks Sep 8, 2026 Can a VLM annotate a robot run? Results on a revised WGO-Bench Samuel Alexander, Sankalp Nagaonkar Agents Aug 9, 2026 Claude edited our launch video with VideoDB skills Ashutosh Trivedi Agents Aug 9, 2026 Giving an agent a YouTube video it can search Ashutosh Trivedi Engineering Aug 9, 2026 Twelve Labs alternatives, compared Ashutosh Trivedi Engineering Aug 9, 2026 Video RAG: how it works Ashutosh Trivedi Agents Aug 8, 2026 Agentic Streams: an agent researches a topic and streams you the briefing Ashutosh Trivedi Papers Aug 8, 2026 Search over the Visual World: persistent visual memory, layered indexes, and source-grounded evidence Nagaonkar, Garg, Raj, Choithani, Trivedi Agents Aug 7, 2026 Real-time visual perception for agents Ashutosh Trivedi Agents Aug 7, 2026 From an RTSP stream to events an agent can act on Ashutosh Trivedi Hackathons Jul 25, 2026 Unlock the Footage: July 2026 project showcase VideoDB Benchmarks Jul 24, 2026 What video retrieval benchmarks taught us about ground truth Samuel Alexander Essay Jul 13, 2026 DeepSeek’s visual primitives and the missing layer Sankalp Nagaonkar, Ashutosh Trivedi Essay Jul 7, 2026 JEPA, from language models to world models Sankalp Nagaonkar, Ashutosh Trivedi Agents Jun 29, 2026 TinyFish × VideoDB: the web your agent browses, as video it can search Ashutosh Trivedi Hackathons Jun 13, 2026 Agent Forge: June 2026 project showcase VideoDB Essay Jun 11, 2026 Episodic memory for agents Ashutosh Trivedi Essay Jun 9, 2026 Video was built for playback, not perception Ashutosh Trivedi Essay Jun 7, 2026 MP4 is the wrong primitive Ashutosh Trivedi Essay Jun 5, 2026 Perception is the missing layer Ashutosh Trivedi Agents Jun 4, 2026 Bloom: record your screen, search it later Lalit Gupta Engineering Jun 4, 2026 Focusd: a coach for your workday Sankalp Nagaonkar Agents Jun 4, 2026 A camera feed for your OpenClaw agent Ashutosh Trivedi Agents · app Jun 4, 2026 Bloom, AI-ready screen recorderAn agentic Loom alternative where recordings become inputs for AI. Ashutosh Trivedi Agents · app Jun 4, 2026 Focusd, Productivity TrackerScreen memory for understanding where your workday went. Sankalp Nagaonkar Agents · app Jun 4, 2026 OpenClaw MonitoringCCTV for computer-use agents. Ashutosh Trivedi Essay Jun 3, 2026 Infrastructure that sees and edits Ashutosh Trivedi Essay Jun 1, 2026 Why agents are blind today Ashutosh Trivedi Benchmarks May 19, 2026 A strong general-purpose VLM still fails on a chessboard Sankalp Nagaonkar Agents May 19, 2026 Deep Search: finding exact moments in video Sankalp Nagaonkar Agents May 19, 2026 Pair Programmer: live screen and audio context for coding agents Rohit Garg Hackathons May 16, 2026 Global Hackathon: May 2026 project showcase VideoDB Benchmarks May 15, 2026 Evaluating video VLMs on your own task Sankalp Nagaonkar Engineering May 8, 2026 Dispatch #001: playable evidence, thinking tokens, agentic streams Ashutosh Trivedi Agents May 8, 2026 How I built Call.md on VideoDB Om Agents · app May 8, 2026 Call.md, Meeting CopilotTurn meetings into live agent loops. Ashutosh Trivedi Agents · app May 8, 2026 Cool QR Codes StudioA no-login QR code studio with scanability checks. Ashish Agents · app May 8, 2026 DictateLocal-first macOS dictation that pastes speech into any app. Ashish Engineering Apr 20, 2026 Reuse HTTP connections in Python services Rohit Garg Engineering Apr 17, 2026 Caching CORS preflights: hundreds of milliseconds per call Om Gate Papers Apr 13, 2026 Do thought streams matter? Evaluating reasoning in Gemini vision-language models for video scene understanding Sharma, Nagaonkar, Choithani, Trivedi Engineering Jan 2, 2026 A Postgres backup is not real until you restore it Rohit Garg Engineering Sep 17, 2025 The 6 MB Lambda limit: compress before, not after Rohit Garg Engineering Aug 25, 2025 Searching the slides in a conference recording Ashutosh Trivedi Benchmarks Aug 25, 2025 NFL game analysis: 80% fewer VLM hallucinations Sankalp Nagaonkar Engineering Aug 25, 2025 Smaller Python ML containers Lalit Gupta Engineering Aug 13, 2025 VideoDB × TwelveLabs: search across your video library Ashutosh Trivedi Engineering Jun 11, 2025 VideoDB × LlamaIndex: video in your RAG pipeline Ashutosh Trivedi Papers Feb 10, 2025 Benchmarking vision-language models on optical character recognition in dynamic video environments Nagaonkar, Sharma, Choithani, Trivedi
Nothing here yet.