Supriya Adapa
Here.
Data × Design.

# Object Detection — YOLOv3 + OpenCV
# Author: Your Name | JNTUK Bachelor's Project

import cv2, numpy as np

CONF = 0.50   # confidence threshold
NMS  = 0.40   # non-max suppression
SIZE = 416    # YOLO input size

CLASSES = ["person","bicycle","car","motorbike",
           "cat","dog","chair","book","laptop"
           ... 71 more COCO classes]

def detect(image_path):
    net = cv2.dnn.readNetFromDarknet(
        "yolov3.cfg", "yolov3.weights")
    img = cv2.imread(image_path)
    h, w = img.shape[:2]

    blob = cv2.dnn.blobFromImage(
        img, 1/255.0, (SIZE,SIZE), swapRB=True)
    net.setInput(blob)
    outputs = net.forward(get_output_layers(net))

    boxes, confs, ids = [], [], []
    for out in outputs:
        for det in out:
            scores = det[5:]
            cid    = int(np.argmax(scores))
            conf   = float(scores[cid])
            if conf > CONF:
                cx, cy = int(det[0]*w), int(det[1]*h)
                bw, bh = int(det[2]*w), int(det[3]*h)
                boxes.append([cx-bw//2,cy-bh//2,bw,bh])
                confs.append(conf)
                ids.append(cid)

    nms = cv2.dnn.NMSBoxes(boxes,confs,CONF,NMS)
    for i in nms.flatten():
        x,y,bw,bh = boxes[i]
        label = f"{CLASSES[ids[i]]} {confs[i]:.0%}"
        cv2.rectangle(img,(x,y),(x+bw,y+bh),color,2)
        cv2.putText(img,label,(x,y-5),...)
    cv2.imwrite("result.jpg", img)

# Highway Vehicle Detection & Speed Estimation
# Author: Your Name | NJIT Image Processing Course

import cv2, numpy as np
from collections import defaultdict

MIN_AREA = 1200   # minimum contour area (px²)
SL1_Y   = 0.45   # speed line 1 position
SL2_Y   = 0.65   # speed line 2 position
DIST_M  = 5.0    # real-world distance between lines

def classify_vehicle(w, h):
    if   w*h > 25000 and w/h > 1.8:
        return "TRUCK", (0,180,255)
    elif w*h > 8000:
        return "CAR", (0,230,80)
    else:
        return "MOTO", (255,140,0)

def process_video(path):
    cap  = cv2.VideoCapture(path)
    bg   = cv2.createBackgroundSubtractorMOG2(
        history=200, varThreshold=50)
    count = 0
    speeds, cross_t = {}, {}

    while True:
        ret, frame = cap.read()
        if not ret: break
        H,W = frame.shape[:2]

                # Background subtraction → contours
        mask = bg.apply(frame)
        mask = cv2.morphologyEx(mask, cv2.MORPH_CLOSE, kernel)
        cnts,_ = cv2.findContours(
            mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

        for c in cnts:
            if cv2.contourArea(c) < MIN_AREA: continue
            x,y,w,h = cv2.boundingRect(c)
            vtype,col = classify_vehicle(w,h)
            cv2.rectangle(frame,(x,y),(x+w,y+h),col,2)
    return count

-- Cloud Connector — Multi-Tenant Schema
-- Author: Your Name | Arete IT Services
-- Impact: +35% sales after deployment

CREATE TABLE tenants (
    tenant_id   INT IDENTITY(1,1) PRIMARY KEY,
    tenant_code VARCHAR(50)  UNIQUE NOT NULL,
    plan_tier   VARCHAR(20)  DEFAULT 'starter',
    is_active   BIT           DEFAULT 1
);

CREATE TABLE cloud_resources (
    resource_id   INT IDENTITY(1,1) PRIMARY KEY,
    tenant_id     INT NOT NULL REFERENCES
                  tenants(tenant_id),
    resource_type VARCHAR(50),
    status        VARCHAR(20) DEFAULT 'running',
    cost_per_hour DECIMAL(10,6)
);

-- Row-Level Security: tenants only see their rows
CREATE FUNCTION dbo.fn_TenantPredicate
    (@tenant_id INT)
RETURNS TABLE WITH SCHEMABINDING AS
    RETURN SELECT 1 AS result
    WHERE @tenant_id = CAST(
        SESSION_CONTEXT(N'tenant_id') AS INT);

CREATE SECURITY POLICY CloudResourcesPolicy
    ADD FILTER PREDICATE
    dbo.fn_TenantPredicate(tenant_id)
    ON dbo.cloud_resources;

-- Stored proc: set tenant context on login
CREATE PROCEDURE dbo.sp_TenantLogin
    @tenant_code VARCHAR(50)
AS BEGIN
    DECLARE @tid INT
    SELECT @tid = tenant_id
    FROM   tenants
    WHERE  tenant_code = @tenant_code
    EXEC sp_set_session_context
         N'tenant_id', @tid
END

// YuktiAI — Agent Configuration
// 6 AI agents running the entire business

const agents = [
  {
    id:    "vikram-bda",
    name:  "Vikram — Business Development",
    model: "deepseek/deepseek-r1:free",
    tools: ["apollo_scrape", "lead_score", "supabase_insert"],
    schedule: "0 9 * * 1-5" // 9am weekdays
  },
  {
    id:    "arya-soa",
    name:  "Arya — Sales Outreach",
    model: "google/gemini-flash-1.5:free",
    tools: ["gmail_send", "calendly_book", "ab_test"],
    sequence: [0, 3, 7, 10, 14] // email days
  },
  {
    id:    "priya-rtsa",
    name:  "Priya — Support Agent",
    model: "anthropic/claude-haiku-4-5",
    tools: ["crisp_chat", "vector_search"],
    sla_hours: 2
  }
];

async function runAgent(agentId, task) {
  const agent = agents.find(a => a.id === agentId);
  const res   = await fetch("https://openrouter.ai/api/v1/chat/completions", {
    method: "POST",
    body: JSON.stringify({
      model: agent.model,
      messages: [{ role: "user", content: task }]
    })
  });
  return (await res.json()).choices[0].message.content;
}