01 · Halaman today: what the public record shows

Compiled 22 September 2026 from public web sources only (search snippets; the company's own site could not be fetched directly from this environment). Every statement carries its source. Third-party estimates are marked as such. The owner should correct anything that is wrong or out of date.

1. Identity

2. What Halaman makes

All products are paper- and board-based, printed offset or digital, converted in-house (design, prepress/CTP, printing, lamination, die-cutting, folding-gluing, finishing).

Sectors named by the company: textile/apparel (primary), cosmetics, food, pharmaceutical, household chemical (DQS certification scope), and "specialty brands". The history page states Halaman "has become a global nominated supplier of large groups and brands". A certification blog post adds that "about 70% of Halaman's customers inquire about various certifications", which is consistent with a brand-nominated, audit-driven customer base.

The references page shows twelve customer logos (not indexed by search engines; read from the page itself): Lacoste, United Colors of Benetton, Etam, Burger King, Salomon, Under Armour, H&M, Hugo Boss, Tesco, Monoprix, Boots and Le Bourget Paris 1924. Eight are apparel, lingerie, hosiery or sportswear brands, three are retailers with own-brand clothing or beauty, one is quick-service food; five are French. Details in doc 03.

Export markets named on the LinkedIn company page: France, Italy, United Kingdom, Germany, Slovakia and other European countries. The majority of production is exported.

3. Machine park and technology (public)

The machine-park page on halaman.com (extracted 22 September 2026, see research/owner-supplied-page-extraction.md) lists the full fleet:

Two B1 and two B2 Heidelberg presses, three digital presses, six flatbed die-cutters and five folder-gluers is a large mixed fleet for a 200-person plant: scheduling and changeover minimisation across it is a real optimisation problem (doc 02), and the Heidelberg fleet is Prinect-capable, which matters for press data (to verify whether Prinect is licensed).

Not visible anywhere in public: the ERP/MIS, prepress workflow software, CRM, or whether the presses report through Prinect / HP PrintOS. Those are the first questions in docs/09-discovery-questionnaire.md.

4. Certifications and sustainability claims

Not found anywhere: ISO 15378 (pharma packaging), EcoVadis, PEFC.

Compliance flag. The EU Empowering Consumers directive bans generic "climate neutral" claims that rest only on offsetting from 27 September 2026 (ClimatePartner's own page says so). Any customer-facing label or marketing text using that wording for EU customers should be re-worded now (for example to a specific, substantiated claim such as "printed with FSC board, 100% recyclable, emissions calculated and reduced by X%"). This is an easy early task for an AI document-review workflow (see doc 02).

5. Fairs and marketing footprint

• FachPack, Nuremberg (September 2025, via Halaman Packaging GmbH).
• Cosmopack / Cosmoprof Worldwide Bologna (March 2025 and March 2026).
• All4Pack, Paris (2018/2019 edition, per a "thank you for visiting our stand" blog post).
• Blog and LinkedIn posts in EN/TR/DE; a Slovak contact profile suggests an agent or partner in Slovakia (unverified).
• A World Bank Group delegation (MIGA guarantees leadership) visited the factory in July 2025. A World Bank blog states that World Bank Group guarantees channelled through Türk Eximbank financed Halaman's digital press investment and describes Halaman as a "family-owned printer of packaging for some of Europe's best-known brands" (see research/customers.md). The German site says Halaman exports to over 30 countries.
• Not found: Luxe Pack, interpack, Texworld, Première Vision, packaging awards, exporter rankings.

6. Timeline

7. What this means for an AI programme (bridge to doc 02)

1. The plant is already digital where it matters. Two HP Indigo presses, Scodix, a KAMA die-cutter and a camera inspection unit all generate machine data. HP PrintOS and Heidelberg Prinect (if licensed) expose job, uptime and waste data that scheduling and quality AI can use without new sensors.
2. The commercial side is where public evidence is thinnest. No CRM, MIS or portal is visible; new business appears to come from fairs, nominations and referrals. Quoting speed, RFQ handling and outbound prospecting are therefore the highest-leverage AI targets (docs 02, 04, 05, 06).
3. Sustainability is a real differentiator but the claims need updating for the EU rules (EmpCo, PPWR, EUDR); the data collection behind them is a natural automation task (doc 07).
4. Two-language operation (TR factory, DE office) plus EN/FR/IT customers makes multilingual assistants for sales and customer service unusually valuable.
5. The core customer cluster is next door. Türkiye is the world's second-largest sock exporter (USD 1.16 billion in 2025, 78% of the 1,034 firms in Istanbul), and at least one large exporter, Şimşek Çorap, sits in the same Akçaburgaz neighbourhood as the Halaman plant (see research/market-textile.md). Header cards and paper hangers for this cluster are the most natural prospecting ground.