Competitor Review Mining

import os, requests, time
from collections import defaultdict

G2 = os.environ["G2_API_KEY"]
MV = os.environ["MAVERA_API_KEY"]
G2_BASE = "https://data.g2.com/api/v1"
G2_H = {"Authorization": f"Token token={G2}", "Content-Type": "application/vnd.api+json"}
MV_H = {"Authorization": f"Bearer {MV}", "Content-Type": "application/json"}

COMPETITORS = [
    {"name": "CompetitorA", "product_id": "competitor-a-product-id"},
    {"name": "CompetitorB", "product_id": "competitor-b-product-id"},
]

all_intel = []
for comp in COMPETITORS:
    reviews = []
    page = 1
    while len(reviews) < 100:
        r = requests.get(f"{G2_BASE}/survey-responses",
            headers=G2_H,
            params={
                "filter[product_id]": comp["product_id"],
                "page[size]": 50,
                "page[number]": page,
            })
        if r.status_code == 429:
            time.sleep(1)
            continue
        r.raise_for_status()
        data = r.json().get("data", [])
        if not data:
            break
        reviews.extend(data)
        page += 1
        time.sleep(0.1)

    loves, hates, recs = [], [], []
    stars = []
    for rev in reviews:
        attrs = rev.get("attributes", {})
        stars.append(attrs.get("star_rating", 0))
        for key, val in attrs.get("comment_answers", {}).items():
            text = val if isinstance(val, str) else val.get("text", "")
            k = key.lower()
            if "love" in k or "best" in k:
                loves.append(text[:200])
            elif "dislike" in k or "hate" in k:
                hates.append(text[:200])
            elif "recommend" in k:
                recs.append(text[:200])

    avg = sum(stars) / len(stars) if stars else 0
    all_intel.append({
        "name": comp["name"],
        "review_count": len(reviews),
        "avg_rating": round(avg, 1),
        "top_loves": loves[:10],
        "top_hates": hates[:10],
        "recommendations": recs[:5],
    })

# Build analysis prompt
intel_block = ""
for ci in all_intel:
    intel_block += f"\n## {ci['name']} ({ci['review_count']} reviews, avg {ci['avg_rating']}/5)\n"
    intel_block += f"LOVE:\n" + "\n".join(f"- {l}" for l in ci["top_loves"][:5]) + "\n"
    intel_block += f"HATE:\n" + "\n".join(f"- {h}" for h in ci["top_hates"][:5]) + "\n"
    intel_block += f"RECOMMENDATIONS:\n" + "\n".join(f"- {r}" for r in ci["recommendations"][:3]) + "\n"

analysis = requests.post("https://app.mavera.io/api/v1/mave/chat",
    headers=MV_H,
    json={"message": f"""Analyze these G2 competitor reviews and identify positioning opportunities for us.

{intel_block}

Produce:
1. Each competitor's core strengths (where they're hard to beat)
2. Each competitor's vulnerabilities (where their users are frustrated)
3. Positioning opportunities (where we can win based on their weaknesses)
4. Messaging angles (specific claims we can make that competitors can't)
5. Feature gaps (what their users want that we could build/highlight)
6. Risk areas (where competitors are improving — watch list)"""}).json()

print("=== Competitive Intelligence from G2 ===")
print(analysis.get("content", "")[:2000])

const G2 = process.env.G2_API_KEY;
const MV = process.env.MAVERA_API_KEY;
const G2_BASE = "https://data.g2.com/api/v1";
const G2_H = { Authorization: `Token token=${G2}`, "Content-Type": "application/vnd.api+json" };
const MV_H = { Authorization: `Bearer ${MV}`, "Content-Type": "application/json" };

const COMPETITORS = [
  { name: "CompetitorA", productId: "competitor-a-product-id" },
  { name: "CompetitorB", productId: "competitor-b-product-id" },
];

const allIntel = [];
for (const comp of COMPETITORS) {
  const reviews = [];
  let page = 1;
  while (reviews.length < 100) {
    const res = await fetch(
      `${G2_BASE}/survey-responses?filter[product_id]=${comp.productId}&page[size]=50&page[number]=${page}`,
      { headers: G2_H }
    );
    if (res.status === 429) { await new Promise((r) => setTimeout(r, 1000)); continue; }
    if (!res.ok) throw new Error(`G2 ${res.status}`);
    const data = (await res.json()).data || [];
    if (!data.length) break;
    reviews.push(...data);
    page++;
    await new Promise((r) => setTimeout(r, 100));
  }

  const loves = [], hates = [], recs = [], stars = [];
  for (const rev of reviews) {
    const attrs = rev.attributes || {};
    stars.push(attrs.star_rating || 0);
    for (const [key, val] of Object.entries(attrs.comment_answers || {})) {
      const text = typeof val === "string" ? val : val?.text || "";
      const k = key.toLowerCase();
      if (k.includes("love") || k.includes("best")) loves.push(text.slice(0, 200));
      else if (k.includes("dislike") || k.includes("hate")) hates.push(text.slice(0, 200));
      else if (k.includes("recommend")) recs.push(text.slice(0, 200));
    }
  }

  allIntel.push({
    name: comp.name, count: reviews.length,
    avg: stars.length ? +(stars.reduce((s, v) => s + v, 0) / stars.length).toFixed(1) : 0,
    loves: loves.slice(0, 10), hates: hates.slice(0, 10), recs: recs.slice(0, 5),
  });
}

const intelBlock = allIntel.map((ci) =>
  `## ${ci.name} (${ci.count} reviews, avg ${ci.avg}/5)\nLOVE:\n${ci.loves.slice(0, 5).map((l) => `- ${l}`).join("\n")}\nHATE:\n${ci.hates.slice(0, 5).map((h) => `- ${h}`).join("\n")}\nRECS:\n${ci.recs.slice(0, 3).map((r) => `- ${r}`).join("\n")}`
).join("\n\n");

const analysis = await fetch("https://app.mavera.io/api/v1/mave/chat", {
  method: "POST", headers: MV_H,
  body: JSON.stringify({
    message: `Analyze G2 competitor reviews:\n\n${intelBlock}\n\nProduce: 1) Strengths 2) Vulnerabilities 3) Positioning opportunities 4) Messaging angles 5) Feature gaps 6) Risk watch list`,
  }),
}).then((r) => r.json());

console.log("=== Competitive Intelligence ===");
console.log((analysis.content || "").slice(0, 2000));