You will get AI Video Analytics (YOLO + OpenCV)
Project details
Add real‑time object detection to any camera feed using YOLOv8 and OpenCV. I integrate GPU‑accelerated (TensorRT) or CPU‑based inference with your existing dashboard or build a new one. Detect people, vehicles, packages, or custom classes. Alerts go to email, Slack, or webhook. Optimized for low latency on live RTSP/WebRTC streams. Source code included in Pro and Enterprise tiers.
What's included
| Service Tiers |
Starter
$400
|
Standard
$900
|
Advanced
$1,600
|
|---|---|---|---|
| Delivery Time | 5 days | 8 days | 12 days |
Number of Revisions | 1 | 2 | 3 |
Number of Pages | 1 | 1 | 1 |
Design Customization | - | - | - |
Content Upload | - | - | - |
Responsive Design | - | - | - |
Source Code | - |
Optional add-ons
You can add these on the next page.
Additional Revision
+$50
Extra Camera Feed
+$100
Custom Dashboard (React)
(+ 2 Days)
+$500
Face Recognition
(+ 2 Days)
+$350Frequently asked questions
About Regan
C++, FFmpeg, WebRTC, AI Voice Agent, Twilio, Python, SIP, VoIP, 3CX
Iligan, Philippines - 6:02 pm local time
By combining deep media server engineering (GStreamer, FFmpeg, Janus, LiveKit) with AI agent frameworks and telecom protocols (SIP, RTP, WebRTC), I create solutions where a single platform handles live video ingest, AI‑driven voice interactions, and PSTN/SIP connectivity - often cutting cloud costs by 40‑70% through on‑prem and edge processing.
🔹 What I Deliver – Core Service Lines
✅ AI Voice Agents & Telephony
- LLM‑powered voicebots connected to live calls via SIP trunks, Twilio, or Plivo.
- Real‑time speech‑to‑text / text‑to‑speech (Deepgram, ElevenLabs, PlayHT) integrated into media servers.
- Conversational AI workflows (Vapi, Retell, custom Node.js/Go) with interruption handling and emotion detection.
- SIP‑to‑WebRTC gateways, SIP trunking, and PBX integration – including deep 3CX setup, customization, and migration alongside Asterisk, FreeSWITCH, and Kamailio.
✅ Low Latency Video Streaming
- WebRTC (SFU/MCU), SRT, LL‑HLS, Media over QUIC (MoQ) – sub‑second glass‑to‑glass.
- Custom media servers (Janus, mediasoup, LiveKit, custom C++/Rust) scaling to 10K+ viewers.
- OBS Studio integrations with hardware AV1/HEVC encoding and WebRTC output.
- OTT sports, live events, and interactive broadcasts.
✅ AI‑Driven Video Analytics
- Real‑time object detection, tracking, face recognition (YOLOv8/v11, OpenCV, MediaPipe).
- Edge deployment on Jetson, Raspberry Pi, or NPU‑accelerated hardware (TensorRT, ONNX).
- Integration with AI voice layer for alerting, commentary, or conversational queries over video feeds.
✅ Embedded & IoT Streams
- C++/Rust firmware for IP cameras (RTSP, ONVIF, MQTT) on Yocto/Buildroot.
- GStreamer pipelines on resource‑constrained devices, hardware‑accelerated encoding.
- Fleet management via MQTT, CoAP, gRPC, Modbus.
🔹 Technical Toolkit
- Languages: C++, Python, Go, Rust, TypeScript/Node.js, SQL
- Protocols: SIP, RTP, RTCP, WebRTC, SRT, RTMP, MoQ, LL‑HLS, LL‑DASH, RTSP, ONVIF, MQTT, CoAP
- Voice & AI Agents: Twilio, Plivo, 3CX, Asterisk, FreeSWITCH, Kamailio, Vapi, Retell, Deepgram, ElevenLabs, OpenAI/LLMs, Rasa
- Streaming & Media: FFmpeg, GStreamer, Janus, mediasoup, LiveKit, Wowza, OBS Studio
- AI/ML & Optimization: YOLO, TensorRT, ONNX Runtime, OpenVINO, OpenCV, MediaPipe
- Cloud & Infra: AWS, GCP, Docker, Kubernetes, Prometheus, Grafana, SIP trunk providers
- Frontend/Backend: React, Next.js, FastAPI, Node.js, Redis, PostgreSQL
🔹 How I Work
Architecture audits → multi‑modal system design → prototype → production. I prioritize clean observability, HIPAA‑ready compliance when required, and thorough documentation.
🔹 Why Clients Choose Me
✅ End‑to‑end video‑plus‑voice architectures, not isolated features
✅ Sub‑second latency for both streaming and conversational AI
✅ Hybrid edge/cloud designs that slash egress and compute costs
✅ Clear daily communication, reliable delivery, and architecture you can scale on
📌 Available for: Long‑term contracts, greenfield MVPs, architecture reviews, and team training.
📩 Message me your project – I respond within 4 hours.
Steps for completing your project
After purchasing the project, send requirements so Regan can start the project.
Delivery time starts when Regan receives requirements from you.
Regan works on your project following the steps below.
Revisions may occur after the delivery date.
Step 2
I review your camera feeds and detection list. Within 24 hours, I provide a test video with bounding boxes for your approval.
Step 3
I integrate the model into your pipeline – supports RTSP, HLS, or uploaded files. Outputs JSON, WebSocket, or a simple dashboard.