reelbot
A pipeline that turns one prompt into a finished 30 to 45 second reel: AI script, AI voice, Remotion video and a scheduled post.
AI AGENTS AND AUTOMATION
An AI automation developer connects language models to your tools so repetitive work runs on its own. Salman Faris is an AI automation developer in Kerala whose own pipeline writes, voices, renders and schedules a daily reel. He builds the same kind of workflow for businesses, with review steps where they matter.
For teams losing hours to repetitive tasks that follow clear rules.
Automation works best on jobs that happen often and follow a pattern: reports, follow-ups, content drafts, moving data between tools. AI adds the steps that used to need a person, such as reading a message or writing a first draft.
I run this kind of system myself. reelbot produces a finished reel from one prompt, and Buddy decides which AI model should handle each task. Those projects taught me where automation fails, and how to add checks so a mistake gets caught before it reaches a customer.
Copying data between tools and writing the same updates takes hours.
Enquiries wait because nobody was reminded.
Regular posting is the first thing dropped when work gets busy.
Agreed in writing before work starts.
Connecting forms, sheets, email, WhatsApp and internal tools.
Task-focused agents with clear tools, limits and logs.
Script, voice, video and scheduling, with a draft mode for review.
Daily or weekly summaries generated from your own data.
Real work where it exists, and clearly labelled concepts where it does not.
A pipeline that turns one prompt into a finished 30 to 45 second reel: AI script, AI voice, Remotion video and a scheduled post.
A desktop launcher that runs any AI from one place: Claude Code, Codex, OpenCode or local models, with history saved for every task.
We write down the steps as they happen today.
Anything customer-facing or costly gets a review step.
One complete path first, so the value is real and measurable.
You can see what ran, and you hear about failures.
I build from Calicut for businesses in Kerala and remote clients. Many local teams run on WhatsApp and spreadsheets, so automation starts from those tools rather than asking everyone to change how they work.
Common uses are drafting replies, sorting enquiries, summarising reports and producing content from a template. The best first project is a task that happens often, follows clear rules and is cheap to check.
An AI agent is a program that uses a language model to decide its next step and act through tools, such as searching or updating a record. Good agents have tight limits and logs.
Every workflow is designed with that in mind. Risky steps go through human review, each run is logged and you can pause the system at any time.
For businesses that want AI to do a specific job on their site, not AI for the sake of it.
For creators and brands that need more video without hiring a full team.
For small businesses, startups and teams that need custom software, not a template.
NEXT STEP
Describe one repetitive task and how it is done today.