Update app.py

app.py

@@ -4,7 +4,6 @@ import pandas as pd
 import plotly.express as px
 from io import StringIO
 from huggingface_hub import InferenceClient
-from openai import OpenAI
 
 # --------------------------
 # Streamlit Setup
@@ -17,55 +16,58 @@ st.set_page_config(
 st.title("🧭 Team Career Progression Assistant")
 st.caption("Program Managers • Scrum Masters • People Leads")
 
-
 # --------------------------
-# Global OpenAI client
+# HuggingFace Model (Free)
 # --------------------------
-def get_client():
-    api_key = os.getenv("OPENAI_API_KEY")
-    if not api_key:
-        st.error("Missing OPENAI_API_KEY in Hugging Face Secrets.")
-        return None
-    return OpenAI(api_key=api_key)
+@st.cache_resource
+def load_hf_model():
+    return InferenceClient("google/flan-t5-large")  # strong + free
 
-client = get_client()
+hf_client = load_hf_model()
 
 
 # --------------------------
-# Auto-Generate Next Role (LLM)
+# Auto-Generate Next Role (HF)
 # --------------------------
 def generate_next_role(current_role: str) -> str:
-    if client is None:
-        return "Next Level Role"
-
-    prompt = f"""
-    What is the most likely next-level career role for someone currently working as a '{current_role}'?
-
-    Provide ONLY the job title without explanation.
-    """
+    prompt = (
+        f"Suggest the most likely next-level job title for someone working as a {current_role}. "
+        "Return only the job title, no explanation."
+    )
 
     try:
-        response = client.responses.create(
-            model="gpt-4.1-mini",
-            input=prompt,
-            max_output_tokens=20
-        )
-        return response.output_text.strip()
-
+        result = hf_client.text_generation(prompt, max_new_tokens=30)
+        return result.strip()
     except Exception as e:
-        st.error(f"Next Role LLM Error: {e}")
+        st.error(f"Next Role Generation Error: {e}")
         return "Next Level Role"
 
 
 # =======================
-# OpenAI LLM – 30-60-90 Plan
+# Generate 30-60-90 Plan (HF)
 # =======================
-def generate_llm_plan(role, next_role):
-    prompt = f"Create a 30-60-90 plan for a {role} moving to {next_role}."
+def generate_llm_plan(role: str, next_role: str) -> str:
+    prompt = f"""
+Create a detailed professional 30-60-90 Day Plan for someone currently working as a {role},
+who is preparing to grow into the next-level role: {next_role}.
+
+The response must follow this structure:
+
+### 30-Day Goals
+- ...
+
+### 60-Day Goals
+- ...
+
+### 90-Day Goals
+- ...
+"""
+
     try:
-        return hf_client.text_generation(prompt, max_new_tokens=180)
+        result = hf_client.text_generation(prompt, max_new_tokens=350)
+        return result.strip()
     except Exception as e:
-        st.error(f"HF Plan Error: {e}")
+        st.error(f"30-60-90 Plan Error: {e}")
         return "Could not generate plan."
 
 
@@ -129,7 +131,7 @@ with st.sidebar:
     file = st.file_uploader("Upload CSV", type=["csv"])
     use_sample = st.checkbox("Use sample data", value=True)
 
-    st.header("🤖 LLM Options")
+    st.header("🤖 AI Options")
     use_llm = st.checkbox("Enable AI-generated 30-60-90 plans", value=False)
 
 
@@ -155,8 +157,9 @@ df["ReadinessScore"] = df.apply(score, axis=1)
 df["SuggestedNextRole"] = df.apply(suggest_next_role, axis=1)
 df["RecommendedActions"] = df.apply(suggest_actions, axis=1)
 
-# LLM Mode
+# AI Mode
 if use_llm:
+    st.info("Generating 30-60-90 plans using FREE HuggingFace models...")
     df["LLMPlan"] = df.apply(
         lambda row: generate_llm_plan(
             row["CurrentRole"],