Concall search — real questions, real answers

“Which companies have spoken about margin expansion?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=margin+expansion&limit=10"

The signal clusters in healthcare (BIOCON: “transition underway to sustainable growth, margin expansion and cashflow led value-creation”; FORTIS: “delivered margin expansion compared to the corresponding quarter last year”; ASTERDM at 21% EBITDA margin vs 20% Q3FY25) and EMS / industrial (SYRMA at 11% EBITDA margin, +78% YoY; JYOTICNC at 26.8% EBITDA margin, +37% growth).

Premium consumer shows up too — ABDL on “40%+ gross margins, highest in the mass premium category”, INDHOTEL with 39-40% segment EBITDA. Notably absent: industrial commodities, mass FMCG. Pivot: q=margin+expansion&symbol=<company> for that one company's full narrative across quarters.

“Where are the heaviest planned capex commitments right now?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=capex+guidance&limit=6"

POWERGRID dominates with the “multi-year capex & capitalization runway” deck — transmission build-out tied to renewable evacuation. TEGA calls out a Chile project on track for Q2FY27 commercial production. ACUTAAS points to ₹143 cr 9-month capex for Jhagadia battery chemicals + pilot plant. ABDL announces Phase 2 incremental capex (Jan 2026) across critical value-chain nodes.

The pattern: discrete sectoral surges (transmission grid, battery chemicals, alcohol distilling), not broad-based corporate capex. LLOYDSME discussion frames it Q3 + 9M FY26 as their strongest standalone quarters — cyclical steel capex returning.

“Is rural demand actually coming back? Who's saying so?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=rural+demand+revival&limit=5"

Yes, but reluctantly and tied to GST. EMAMILTD: “rural, which was not performing, we are seeing some growth coming in from rural areas. With this reduction in GST rates, we are very hopeful…”. BOSCHLTD: “strong rural interest for two-wheelers… September GST reforms significantly boosted”. TVSMOTOR notes growth was 0.6% rural vs 3.6% urban — thin, but they thank the GST cut.

So the “revival” is a single-digit pickup catalysed by a tax change, not a structural demand turn. PIDILITIND is being asked the question by analysts — the buy-side wants confirmation that hasn't fully arrived yet.

“Who's deploying real money into AI / data-center infra in India?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=data+center+demand&limit=5"
curl -sH "$H" "https://genka.dev/latest/concalls/search?q=AI+capex&limit=5"

Three distinct stories. ANANTRAJ is the pure-play DC-REIT bet — Delhi-NCR land bank with 6 MW operational + cloud services on ~320 acres of debt-free land (top hits at score 9.4 and 7.7). TCS ran a fireside chat 2025-10-31 explicitly titled “TCS' AI data center investment and opportunity” — large-cap IT building captive infra. CUMMINSIND cites a meaningful data-center revenue contribution of ₹1,271 cr that quarter — the genset / power-backup angle.

Cross-reference into transmission via POWERGRID (top hit on q=AI+capex) — the bottleneck for Indian DC scale is grid-side, not server-side.

“How exposed are Indian companies to Chinese end-demand?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=China+demand&limit=5"

The corpus answer is mostly steel. SAIL repeatedly cites World Steel Association data showing China dominating global production despite reductions in its own consumption — the spillover risk to Indian steel pricing. HEG (graphite electrodes) frames it as an EAF (electric arc furnace) story — ex-China EAF share rose from 44% to 51% (2015-2024) while China's stays at 11%, driving graphite electrode demand outside China. HEG also flags China crude steel output -2.6% YoY due to weak construction.

Pivot: q=Chinese+steel+dumping or q=anti-dumping+duty to find the trade-policy commentary.

“What was the GST 2026 transition impact on FMCG and others?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=GST+rate+change&limit=5"

Universal short-term hit, signs of medium-term tailwind. HINDUNILVR: “quarter reflected a transitory impact of GST changes and prolonged monsoon… EBITDA margin at 23.2% was lower by 90 bps”. DABUR: “temporary disruptions post GST reforms announcement… GST rate reductions in ~66% of the India portfolio likely to boost consumption”. GODREJCP echoes “UVG performance impacted by GST transition in India”.

Cross-sector spread: HDFCLIFE separately on competitive dynamics post-rate-cycle, LEMONTREE calls out renovation + tech + GST impact at 6.4% of revenue this quarter. The shape: 1-2 quarter margin compression for distributors, then volume tailwind from price elasticity.

“Is anyone talking about channel destocking still?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=channel+inventory+destocking&limit=5"

Yes, in three places. BLUEJET healthcare (Feb 2026): “there is clearly some channel inventory for which there is some destocking happening” — pharma intermediates. HAVELLS: “channel stocking for cooling products was impacted by milder start to season” — weather-driven, not demand-driven. RRKABEL on “continued caution in discretionary demand and channel level adjustments”. EMAMILTD + KAJARIACER round out with FMCG and tile-distribution context.

Read: discrete pockets, not broad. The structural destocking wave (2024 FMCG) is over; what remains is event-driven (weather, inventory cycle).

“Who's calling out a clean deleveraging plan?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=deleveraging+plan&limit=5"

NUVOCO cement: “driving efficiency and deleveraging in FY25 amidst macro challenges… achieved volume of 19.4 mmt”. UPL agrochem on platform-led deleveraging post Arysta + ADIA / TPG infusion. ADANIPOWER goes deepest with year-by-year senior-debt-to-equity reduction — outstanding regulatory income of ~₹29,000 cr realised FY18-24 funding the paydown.

The Adani Power story is the loudest in the corpus — multiple decks track the credit-profile evolution from BBB. If you're screening for “balance sheet repair” theses, that's the cleanest standalone example.

“What were Indian companies actually saying during the 2020 lockdown?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=lockdown+disruption&limit=5"

Real-time, not retrospective. Pre-2024 docs were 0 in the corpus before this year's Cloudflare-Worker historical backfill. HINDUNILVR 2020-09 documents the state-by-state lockdown impact on FMCG distribution. RELIANCE 2020-07 on “Q1 covid related national lockdown”. DEEPAKNTR 2020-05: “revenue growth was impacted due to 10 days of lockdown”.

Useful when you want the actual contemporaneous narrative arc rather than a 2024 commentator's reconstruction.

“Did Indian sugar companies meaningfully shift to ethanol?”

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=ethanol+blending&limit=5"

BALRAMCHIN is the central figure — integrated sugar-ethanol-bioplastic, repeatedly discussed: “no revision in ethanol prices resulted in decline in margins… sugarcane crushing during the quarter was ~8.4% higher at 387.6 lac quintals owing to early start”, plus commentary on US ethanol import scares easing. EIDPARRY: when sugar prices trade below ethanol parity, mix shifts — pricing arbitrage logic stated explicitly. DCMSHRIRAM reports across chemicals + sugar & ethanol + fenesta + farm solutions segments.

The thesis is real but margin-fragile — the answer to “did they shift?” is yes, and the answer to “does it print money consistently?” is no, gated entirely by GoI ethanol price revisions.

“Show me what RELIANCE management has said about capex specifically.”

Symbol-scoped search drops everyone else; you get the full RELIANCE narrative.

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=capex&symbol=RELIANCE&limit=3"

Three RELIANCE-only hits: “capex at 33,000…”, “earnings growth to outpace capex over the next…”, “flat net debt environment and look at the capex versus our cash…”. Reading the three together: the company is signalling capex is plateauing and cash conversion is improving — a multi-quarter shift.

“Show me only the slide decks — transcripts have too much filler.”

kind=IP filters to investor-presentation slide decks. They have order-book numbers + segment EBITDA tables that don't appear in the Q&A transcript.

curl -sH "$H" "https://genka.dev/latest/concalls/search?q=order+book&kind=IP&limit=5"

Top results: RITES (10.3, IP), THERMAX (8.0), ZENTEC (7.4). Order-book disclosure is densest in slides because that's what management chooses to highlight visually. Pivot to kind=TXN for the same query → you'd get analyst Q&A challenging the order-book quality.

“I have a fragment id — pull the full transcript.”

curl -sH "$H" "https://genka.dev/latest/concalls/8592" | jq '.data | {symbol, disclosure_date, kind, text_len: (.text | length)}'

Returns {symbol, kind, disclosure_date, quarter, text}. Bodies range from 5KB (slide decks) to 80KB (full Q&A transcripts). Useful when you've got a snippet hit you want to read in context, or when you're feeding a single fund's last 4 transcripts into your own LLM.

“Just answer the question — I want a synthesised answer with citations, not a hit list.”

curl -sH "$H" -X POST -H "Content-Type: application/json" \
  -d '{"question":"What did RELIANCE management say about FY26 capex guidance?","symbol":"RELIANCE","limit":3}' \
  https://genka.dev/latest/concalls/answer

Returns {answer, citations[]}. Each citation carries an id so [N] markers in answer dereference back via /concalls/{id}. LLM tier (~1.2× upstream cost). Use this when you want one paragraph instead of five snippets — for everything else, plain search is faster and cheaper.