Environment : Multi-objective Reinforcement Learning
Autonomous procurement · 3-task difficulty system · state → action → reward
step 0 Idle
Select task difficulty
Easy task
Obvious optimal decision
Most vendors activePrices near expectedClear winner
Expected score: ~0.85–0.95
Medium task
Trade-offs required
Some denialsCheap-but-slow vs fast-but-costlyNo obvious pick
Expected score: ~0.65–0.82
Hard task
No perfect answer
Most denyAll prices above budgetQuality vs cost conflict
Expected score: ~0.42–0.68
Vendor dynamics:
Procurement parameters
RL state — episode snapshot
Episode step
—
Vendors remaining
—
Deals closed
—
Best price
—
Budget headroom
—
Cumulative reward
0.000
Live action feed
Policy decision trace
Run the agent to see policy decisions.
Total
10
Active
0
Denied
0
Deals
0
#VendorPrice ₹Del.QualityReliabilityMarginStatusRating

Run the agent to generate results.