Illustrative AI use-cases that each watch one domain, ground in its 360, and recommend the next action — the reasoning layer over the cockpit.
Each use-case owns one job — collect receivables, watch commodity cost, warn on project risk, score bids, defend the order book, reconcile contract terms, hold tender pricing, automate design & galvanizing. They read the same governed data the 360s show, surface the live finding, and recommend the move. Illustrative AI capabilities for a T&D EPC — the reasoning layer over the cockpit.
Illustrative for a T&D EPC — every card reads the same governed data its 360 shows: a live finding, the recommended move, and the value at stake.
Finding. ₹336 Cr of receivables is >60 days late; state-utility and international balances (e.g. EETC (Egypt) / International at 260d) stretch the 214-day book while PGCIL pays fast.
Recommends. Score every open invoice by collection risk and sequence a collections sprint + tighter milestone billing — frees ~₹337 Cr toward the 170-day target.
Finding. Firm steel & zinc prices pressure the thin EBITDA margin — it sits 1.6 pts below the 6% target and management has flagged FY27 commodity-cost pressure.
Recommends. Alert on commodity moves, flag fixed-price contracts lacking price-escalation clauses, and time procurement / hedging to protect margin.
Finding. ₹2.09k Cr of in-flight capability revenue is still ramping; Data-Centre GIS Substations lags at 55% capability capture — a schedule & margin risk.
Recommends. Predict slippage from milestone & site data; put Data-Centre on a dated recovery plan before it drags DSO and margin.
Finding. The addressable T&D tender pipeline is deep; Power Grid Corp (PGCIL) alone carries ₹400 Cr of near-term scope (765 kV / GIS · data-centre · international).
Recommends. Score tenders on win-probability, commodity exposure and working-capital drag before bid; prioritise 765 kV / GIS & data-centre scope and avoid value-destructive low bids.
Finding. ₹390 Cr of the order book is at risk of slippage / non-repeat across the next four quarters.
Recommends. Trigger save-plays and repeat-client actions; protect execution slots and lock follow-on awards before the window closes.
Finding. International & state-utility accounts (e.g. EETC (Egypt) / International at 260d) carry conflicting payment / LC terms across project systems.
Recommends. Reconcile term conflicts on the golden record; standardize export-LC / Incoterms and utility milestone terms to protect the receivables book.
Finding. Standalone estimators discount at 4% vs 3% on governed systems — straight margin leakage on fixed-price tenders.
Recommends. Move all bid pricing onto one governed estimator with a single rate book & approvals; protect the thin tender margin.
Finding. International & MENA projects runs lowest at 82% utilization with 20 open reqs; galvanizing is ~98% utilized ahead of the ₹170 Cr expansion.
Recommends. Automate tower/monopole design and optimise Ranjangaon scheduling & galvanizing throughput; convert paid-for fabrication capacity into output.
Grounded in the live cockpit snapshot via Azure OpenAI.