File size: 7,921 Bytes
38336e8 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 | import sys
import os
from typing import Any, Dict, Tuple, Optional
from uuid import uuid4
# Make sure models module can be imported
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from openenv.core.env_server.interfaces import Environment
from openenv.core.env_server.types import State
from models import RevOpsObservation, RevOpsAction, EnrichmentData
from server.data_generator import get_task_data, get_reps, get_icp_criteria
from server.graders import grader_easy, grader_medium, grader_hard
global_last_grader_score = None
class RevOpsEnvironment(Environment):
SUPPORTS_CONCURRENT_SESSIONS: bool = True
def __init__(self, **kwargs):
self._state = State(episode_id=str(uuid4()), step_count=0)
self.task_id = "task_easy" # DEFAULT
self.state_data = {}
def reset(self, seed: Optional[int] = None, episode_id: Optional[str] = None, **kwargs) -> RevOpsObservation:
self._state = State(episode_id=episode_id or str(uuid4()), step_count=0)
task_id = kwargs.get("task_id")
if not task_id and episode_id in ["task_easy", "task_medium", "task_hard"]:
task_id = episode_id
if task_id:
self.task_id = task_id
task_data = get_task_data(self.task_id)
self.state_data = {
"leads": task_data["leads"],
"current_lead_index": 0,
"enrichment_map": task_data["enrichment_map"],
"crm": task_data["crm"],
"reps": get_reps(),
"icp": get_icp_criteria(),
"action_history": [],
"accumulated_reward": 0.0,
"last_feedback": f"New episode started for {self.task_id}.",
"lead_states": {lead.id: {"enriched": False, "scored": False, "crm_checked": False, "merged": False, "flagged": False} for lead in task_data["leads"]},
"grader_score": None
}
obs = self._get_observation()
obs.done = False
obs.reward = 0.0
return obs
def _get_observation(self) -> RevOpsObservation:
idx = self.state_data.get("current_lead_index", 0)
leads = self.state_data.get("leads", [])
if not leads:
return RevOpsObservation()
if idx < len(leads):
lead = leads[idx]
lead_state = self.state_data["lead_states"][lead.id]
enrichment = self.state_data["enrichment_map"][lead.id] if lead_state["enriched"] else EnrichmentData(enriched=False)
crm = self.state_data["crm"] if lead_state["crm_checked"] else self.state_data["crm"].model_copy(update={"existing_accounts": [], "opportunities": []})
else:
lead = leads[-1]
enrichment = EnrichmentData(enriched=False)
crm = self.state_data["crm"]
return RevOpsObservation(
task_id=self.task_id,
lead=lead,
enrichment=enrichment,
crm=crm,
reps=self.state_data.get("reps", []),
icp_criteria=self.state_data.get("icp", ""),
sla_time_remaining_minutes=60,
last_action_feedback=self.state_data.get("last_feedback", "")
)
def step(self, action: RevOpsAction, **kwargs) -> RevOpsObservation: # type: ignore[override]
if "action_history" not in self.state_data:
self.reset(task_id=self.task_id)
self._state.step_count += 1
self.state_data["action_history"].append(action)
idx = self.state_data["current_lead_index"]
is_done = False
reward = 0.0
feedback = ""
if idx >= len(self.state_data["leads"]):
obs = self._get_observation()
obs.done = True
obs.reward = 0.0
obs.metadata = {"message": "Episode is already finished."}
return obs
current_lead = self.state_data["leads"][idx]
lead_state = self.state_data["lead_states"][current_lead.id]
if action.action_type == "enrich_lead":
if not lead_state["enriched"]:
lead_state["enriched"] = True
reward += 0.1
feedback = "Lead enriched successfully."
else:
feedback = "Lead already enriched."
elif action.action_type == "check_crm":
if not lead_state["crm_checked"]:
lead_state["crm_checked"] = True
reward += 0.1
feedback = "CRM checked successfully."
else:
feedback = "CRM already checked."
elif action.action_type == "update_lead_score":
if self.task_id == "task_medium" and not lead_state["enriched"]:
reward -= 0.2
feedback = "Violation: Scored before enrichment."
else:
current_lead.score = action.score
lead_state["scored"] = True
reward += 0.1
feedback = f"Lead score updated to {action.score}."
elif action.action_type == "merge_with_account":
if not lead_state["crm_checked"] and self.task_id == "task_hard" and current_lead.id == "lead_3_cfo":
reward -= 0.5
feedback = "Violation: Merged without checking CRM first."
else:
lead_state["merged"] = True
reward += 0.1
feedback = f"Merged with account {action.account_id}."
elif action.action_type == "flag_reengagement":
lead_state["flagged"] = True
reward += 0.1
feedback = f"Flagged as re-engagement for opportunity {action.opportunity_id}."
elif action.action_type == "route_to_rep":
if self.task_id == "task_hard" and current_lead.id == "lead_3_cfo" and not lead_state["crm_checked"]:
reward -= 0.4
feedback = "Fatal violation: Routed without checking CRM."
else:
reward += 0.1
feedback = f"Lead routed to rep {action.rep_id}. Moving to next lead."
self.state_data["current_lead_index"] += 1
elif action.action_type == "disqualify":
reward += 0.1
feedback = f"Lead disqualified: {action.disqualification_reason}. Moving to next lead."
self.state_data["current_lead_index"] += 1
else:
feedback = f"Action {action.action_type.value} performed."
self.state_data["accumulated_reward"] += reward
self.state_data["last_feedback"] = feedback
if self.state_data["current_lead_index"] >= len(self.state_data["leads"]):
is_done = True
obs = self._get_observation()
info = {"message": feedback}
if is_done:
if self.task_id == "task_easy":
score = grader_easy(self.state_data["action_history"], self.state_data["leads"])
elif self.task_id == "task_medium":
score = grader_medium(self.state_data["action_history"], self.state_data["leads"])
elif self.task_id == "task_hard":
score = grader_hard(self.state_data["action_history"], self.state_data["leads"])
else:
score = 0.0
info["grader_score"] = score
self.state_data["grader_score"] = score
global global_last_grader_score
global_last_grader_score = score
if score >= 0.7:
reward += 0.5
else:
reward -= 0.5
obs.done = is_done
obs.reward = reward
obs.metadata = info
return obs
@property
def state(self) -> State:
return self._state
def get_full_state(self):
return self.state_data
|