Skip to content

HCL Technologies Ltd (HCLTech)

Company Overview

Business Overview

HCLTech (HCL Technologies Ltd) is a global technology company headquartered in Noida, India, listed on the NSE/BSE and the third-largest India-based IT-services company by revenue. It reports three segments: IT and Business Services, Engineering and R&D Services (ERS), and HCLSoftware (its product arm). Employee count was 223,889 as of June 2026; consolidated revenue for the twelve months ending March 2025 was $13.8 billion. [Evidence — secondary source] 171

Two corporate facts matter most for this report:

  • Geometric Ltd acquisition (announced April 2, 2016). Geometric was an Indian PLM/engineering-software and services firm; the deal brought HCL the DFMPro product line and Geometric's engineering-services business. Geometric's US arm survives as the HCLTech subsidiary "Geometric Americas Inc." DFMPro is today sold under the HCLSoftware brand. [Evidence] 1714
  • ASAP Group acquisition (announced July 12, 2023). A 100% equity purchase of the Ingolstadt-headquartered German automotive engineering firm — over 1,600 engineers across nine German locations, serving German OEMs and Tier-1s in autonomous driving, e-mobility and connectivity. Reuters reported the price at ~€251 million. [Evidence] 23

Automotive Business

Automotive is one of HCLTech's named industry verticals. Verified anchors:

  • Volvo Cars (June 12, 2025) — chose HCLTech as one of its strategic suppliers for engineering services, expanding an existing digital-and-PLM relationship into "delivery of end-to-end engineering solutions at scale," served from an automotive Centre of Excellence in Gothenburg plus offshore centres. No AI content and no quantified outcome in the release. [Evidence] 1
  • ASAP Group (above) — German OEM/Tier-1 client base; services in electrics/electronics, software, consulting, testing/validation and vehicle development. [Evidence] 2
  • Analyst positioning: named a Leader in Everest Group's Software-Defined Vehicle (SDV) Engineering Services PEAK Matrix 2025 and in ISG's Automotive and Mobility 2025 assessments. Note both awards are for electronics and software engineering, not mechanical/moulded parts. [Evidence] 8

Engineering Business

HCLTech's ERS unit spans product engineering, digital engineering, and "Digital Design and Manufacturing" (its PLM practice, covering Siemens, PTC and Dassault toolchains — implementation, migration, and engineering delivery on CATIA/NX/ Creo-based programmes). The automotive services page lists body, chassis and powertrain design plus simulation and certified testing labs. [Evidence] 98

Role in Automotive Moulded Parts Workflow

Workflow Stage Relevant? Notes
Requirements Yes (services + XLM.AI) Requirements management/decomposition with GenAI assist7
Industrial Design Services only Staff-augmentation on client CAD
Concept Design Services only As above
CAD Modelling Yes — services Delivery on CATIA/NX/Creo; PLM practice9
Engineering Review Yes — DFMPro Automated rules-based DFM review inside CAD10
Simulation Services only No owned solver of any kind
DFM Core strength — rules DFMPro injection-moulding checks; Rule Manager; DFX Server batch11
Tool Design Partial DFMPro flags "thin steel conditions on mold"; no tool-design product11
Mould Flow No No mould-flow capability at all — a defining gap
Prototype Services only —
Validation Yes — services ASAP testing/validation (E/E-centric)2
Manufacturing Engineering Yes — services Manufacturing engineering services; XLM.AI process-plan generation7
Production Release Yes — services PLM services

What makes HCLTech structurally different among the engineering service providers: it is the ESP that productised DFM. Where competitors deliver DFM as billable review hours, HCLTech sells a shrink-wrapped rules engine that lives inside every major CAD system — plus the services to configure it. [Synthesis]

AI Strategy

Public AI Vision

HCLTech's AI story is enterprise-wide GenAI and agentic AI, packaged as AI Force (plus AI Foundry for data, and advisory/labs offerings). The pitch is productivity across the software lifecycle and IT operations — not mechanical engineering. [Evidence] 46

Investor Statements

  • Launch framing, March 2024: AI Force "accelerates time-to-value by transforming the software development and engineering lifecycle, delivering greater productivity, improved quality and faster release timelines." [Marketing] 4
  • Vijay Guntur (then President of ERS; by 2026 CTO): "This platform is a true game-changer and some of our pilots with clients in the technology and financial services industries have delivered extremely encouraging outcomes." Note the named pilot industries — technology and financial services, not automotive. [Evidence + Marketing] 4

Engineering AI Strategy

Three streams, in decreasing distance from moulded parts: [Synthesis]

  1. AI Force — GenAI/agents for software engineering, data engineering and IT operations. Built on Azure OpenAI at launch; "model agnostic" by 2.0. The LLM is rented, not owned — consistent with every vendor in this study. [Evidence] 45
  2. XLM.AI — GenAI agents inside ALM/PLM platforms (Siemens, PTC, Dassault, Atlassian, IBM): requirements compliance (INCOSE), test-case generation, BOM validation, process-plan generation from engineering data, design-data validation, and "intelligent part search." This is the closest HCLTech AI gets to mechanical engineering — and it operates on lifecycle data and documents, not geometry. [Evidence] 7
  3. DFMPro's stated philosophy — rules remain the bedrock; AI's role is to augment rules (filter false positives, contextualise violations, mine change orders for candidate rules). Published as strategy, not as shipped product. [Evidence] 16

Timeline of AI Evolution

Date Event
Apr 2, 2016 Geometric Ltd acquisition announced — DFMPro joins HCL (pre-AI-era rules product)17
Jul 12, 2023 ASAP Group acquisition announced (~1,600 German automotive engineers)2
Mar 5, 2024 AI Force launched — GenAI for the software development & engineering lifecycle, built on Azure OpenAI4
2024–2025 AI Force integrations announced (GitHub Copilot, Google Gemini, Anthropic Claude 3 on Amazon Bedrock)6
Jun 12, 2025 Volvo Cars names HCLTech a strategic engineering-services supplier1
Feb 2, 2026 DFMPro blog: "Beyond the Hype: Why Rules-Based Systems Are Still the Bedrock of AI-Driven DFM" — the product team's public position on AI in DFM16
Apr 1, 2026 AI Force 2.0 — agentic, model-agnostic; scope still software/data/ITOps5

Products Relevant to Engineering

DFMPro — CAD-integrated Design-for-Manufacturability software. Purpose: "identifies & provides recommendations to resolve potential downstream manufacturing issues" during design, inside the designer's own CAD system (CATIA V5, 3DEXPERIENCE, NX, Creo Parametric, SOLIDWORKS). AI: none — deterministic rule checks against built-in and customer-defined guidelines. Moulded parts: a dedicated injection-moulding module (§5). [Evidence] 1015

DFX platform add-ons — Rule Manager (configure organisation-specific guidelines as rules), DFX Server (batch checking of large assemblies), DFMPro Costing (early cost visibility), DFX Analytics (dashboards over check results). AI: none described on any product page or brochure. [Evidence] 1513

AI Force / AI Force 2.0 — enterprise GenAI + agentic platform. Relevance to moulded parts: none. Modules are AI Force.Software, .Software.Mod, .ITOps, .Data, plus SAP/Oracle/Pega variants. The word "mechanical" does not appear on the platform page. [Evidence] 6

XLM.AI — "AI-Powered Lifecycle Management" over ALM/PLM systems. Relevance: operates on the PLM backbone automotive programmes run on; includes "intelligent part search" and process-plan generation. Its published case studies are healthcare/pharma, not automotive. [Evidence] 7

ER&D services — engineering delivery (design, simulation support, testing) on client toolchains; ASAP adds German OEM vehicle-development and validation capacity (E/E-weighted). [Evidence] 28

AI Capabilities

1. DFMPro injection-moulding checks — rules, not AI. The module validates "uniform wall thickness, recommended rib parameters, appropriate draft angles on core and cavity surfaces, undercuts, thin steel conditions on mold," with textbook parameter ranges (rib thickness 0.4–0.6× wall; rib-base radius 0.25–0.4× wall; draft ~1–1.5° recommended). Customers can configure the supplied rules and "add new design rules requiring very basic programming knowledge." This is exactly the open-textbook DFM rulebook, productised. Honest test: computes from rules/geometry — not AI, and (to its credit) not marketed as AI.18 [Evidence] 11

2. DFMPro "auto-ignore." The Feb 2026 blog advises: "leverage auto-ignore functionality that learns from organizational actions — those of the community or expert reviewers — to reduce noise." This is the only learning-flavoured claim about the shipping product found anywhere. Limitation: no documentation of a trained model; it reads as remembered reviewer dismissals (a feedback heuristic), not machine learning. Treat the word "learns" with caution. [Evidence — thin] 16

3. DFX Analytics. Dashboards over DFM-check data ("visualizes the data with predefined plots"). The blog suggests this data "can help predict" rule drift — but the product page describes no predictive or ML capability. [Evidence] 1316

4. AI Force (Mar 2024) / AI Force 2.0 (Apr 2026). GenAI + agents for software development, modernisation, IT operations and data engineering. Genuinely LLM-based (Azure OpenAI at launch; multi-LLM later) — so it passes the honest-AI test — but its published scope contains no mechanical engineering, CAD, DFM or moulding use case. Claimed impact ("30% faster releases, 45% fewer defects") is unattributed vendor marketing. [Evidence + Marketing] 456

5. XLM.AI. GenAI agents over ALM/PLM: requirements decomposition and INCOSE-compliance checks, AI-generated test cases, BOM/structure validation, process-plan generation, design-data validation, multilingual translation, "intelligent part search," and a self-learning support bot. Claims "40% faster time-to-market, 50% higher productivity, 25% cost savings" — unattributed. Limitation: document/metadata-bound; nothing indicates geometric understanding; case studies are pharma/healthcare. [Evidence + Marketing] 7

6. The blog's AI-for-moulding vision — future tense. For injection moulding specifically, the blog offers: rules give "hard limits on draft angles and uniform wall thickness to ensure part ejection," while "AI may be able to predict complex cooling-induced warpage that simple geometry rules cannot see," and "generically, AI can learn acceptable design standards based on existing organizational designs which have passed checks." These are possibilities, not products — no shipped capability, no dates, no benchmarks. [Evidence — strategy only] 16

Engineering Workflow Contribution

Two distinct contributions, one per business:

As a product company (HCLSoftware): designer models a part in CATIA/NX/ Creo/SOLIDWORKS → DFMPro runs injection-moulding (or machining, sheet-metal…) rule checks inside the CAD session → violations ranked with recommendations → optionally batch-checked at assembly level on DFX Server → results aggregated in DFX Analytics; organisation-specific rules maintained via Rule Manager. [Evidence] 15

As a services company (ERS): staffing and delivery across OEM programmes — requirements, CAD design, simulation support, E/E and software engineering, testing/validation (ASAP), PLM operations — on the client's toolchain (Volvo: digital + PLM expanding to end-to-end engineering). [Evidence] 12

The two converge in engagements where HCL configures DFMPro with a customer's own rules — verified at Cisco, where HCL "partnered … to add Cisco specific design rules" covering "Sheet Metal, Plastic & Die-cast parts" (§7). No AI participates in either flow today. [Evidence] 14

Public Customer Evidence

Case Studies / Testimonials

  • Cisco Systems (plastic parts — the key moulded-parts testimonial). Biren Kothari, NPI Technical Lead, Mechanical, Cisco San Jose, in the DFMPro brochure: "HCL DFMPro product enabled us to digitize the design rules for a quick & timely feedback… we partnered with HCL to add Cisco specific design rules… DFMPro is the only automated tool for Design for Manufacturability (DFM) analysis besides input from manufacturer or analyzing parts manually. We have DFMPro configured for Sheet Metal, Plastic & Die-cast parts." Qualitative only — no numbers. (Brochure published Nov 2024; the same testimonial reportedly appeared in an Apr 2023 edition.) [Evidence] 14
  • Rockwell Automation. Senior tool designer testimonial: DFMPro "significantly helped us to reduce" part-checking time — the site cites a 60% reduction in design checking time. The strongest quantified DFMPro number found. [Evidence — vendor-hosted testimonial] 10
  • Jabil. "With DFMPro, we reduced the scrap, identified issues for improvement" — contract-manufacturer testimonial, unquantified. [Evidence — vendor-hosted] 10
  • Motorola Solutions. Testimonial praising DFMPro and GeomCaliper (a wall-thickness checking tool, also ex-Geometric). [Evidence — vendor-hosted] 10
  • Ford Otosan. Appears in the customer-logo strip of DFMPro's automotive page — the only automotive OEM/JV name attached to DFMPro publicly. Logo only; no case study. [Evidence — logo only] 12
  • Volvo Cars. Strategic engineering-services supplier (June 2025) — solid services evidence, but no AI, no moulded parts, no metrics. [Evidence] 1

Marketing metrics

DFMPro's site claims "20% reduction in rework, 12% reduction in manufacturing cost, 10% improved time to market, 10% increase in engineering productivity," and its automotive page prices engineering change orders at "$5,000 to $7,500 per change." None are attributed to a named customer or study. [Marketing] 1012

The gap. No public story combines HCLTech + automotive + moulded parts + a quantified outcome. The moulded-parts evidence that exists (Cisco) is high-tech electronics; the automotive evidence that exists (Volvo, ASAP) is not moulded-parts-specific; and none of it involves AI. [Evidence of absence]

Technical Architecture (Inferred)

Evidence

  • DFMPro: native add-ins for CATIA V5, 3DEXPERIENCE, NX, Creo, SOLIDWORKS; a "DFMPro knowledge database" of built-in best practices, supplier capability and cost data; PLM/ERP/MES integration; DFX Server for offline batch checks. [Evidence] 15
  • AI Force: launched on Azure OpenAI, "patented," "system agnostic"; 2.0 is model-agnostic with prompts/RAG/agents/governance layers. [Evidence] 45
  • XLM.AI: agents connecting ALM/PLM platforms (Siemens, PTC, Dassault, Atlassian, IBM) on AWS/Azure/GCP. [Evidence] 7

Synthesis

HCLTech runs two disconnected stacks: a 2000s-era geometric rules engine (DFMPro — feature recognition + parameter checks inside CAD) and a 2020s LLM-orchestration platform (AI Force/XLM.AI — rented models + RAG + agents over enterprise data). Nothing public connects them. The blog's "sandwich architecture" (rules foundation → AI filter layer → human decision) is a published design for connecting them. [Synthesis] 16

Inference

If HCLTech wired XLM.AI's agent layer to DFMPro's violation stream and DFX Analytics data, it would have the raw ingredients of the hybrid it describes — historical waivers to train a false-positive filter, ECO/NCR text to mine for candidate rules. That it has published the concept but shipped none of it suggests the usual blocker: DFMPro's customers' outcome data is not HCLTech's to train on. [Inference]

AI Technologies

In one list: LLMs (Azure OpenAI at AI Force launch; later GitHub Copilot, Gemini and Claude integrations — all rented, none owned); RAG and agentic orchestration (AI Force 2.0); GenAI over PLM/ALM data (XLM.AI, incl. intelligent part search and process-plan generation); rules-based geometric feature checking (DFMPro — not AI, and not claimed to be); analytics dashboards (DFX Analytics). No computer vision on parts, no simulation AI, no surrogate models, no geometry ML anywhere. [Evidence + Synthesis] 467

Research Publications

Papers / Patents

  • The AI Force launch release calls it a "patented GenAI platform" — no patent numbers given, and no independent verification was attempted for this report. [Evidence — thin] 4
  • No academic publications or patents connecting HCLTech to machine learning for DFM, moulding, or CAD geometry were found in this research pass. Geometric Ltd historically published in CAD/CAM circles; nothing recent and AI-related surfaced. [Evidence of absence — limited search]

Notes

Contrast with the platform vendors: Dassault and Autodesk hold CAD-ML patents and (in Autodesk's case) open research code. HCLTech's public technical footprint in geometry-ML is nil — its IP is the rule library and CAD integrations. [Synthesis]

Open Source

Near-nil for engineering AI — reported plainly. [Evidence of absence]

No open-source DFM, CAD-ML or moulding-related repositories associated with HCLTech or the DFMPro team were found. HCLTech's open-source presence relates to its IT/software business, not engineering AI. (Not exhaustively searched; kept in Appendix C.) [Evidence of absence — limited search]

Engineering Service / Platform Mapping

  • CATIA V5 / 3DEXPERIENCE — DFMPro add-in; PLM services practice. [Evidence] 109
  • NX / Teamcenter — DFMPro for NX; Siemens-platform services. [Evidence] 10
  • Creo / Windchill — DFMPro for Creo; PTC-platform services. [Evidence] 10
  • SOLIDWORKS — DFMPro add-in. [Evidence] 10
  • ALM (Atlassian, IBM) + PLM — XLM.AI agents. [Evidence] 7
  • Cloud/LLM — Azure OpenAI, AWS Bedrock (Claude), Google Gemini. [Evidence] 6

The multi-CAD posture is the strategic point: DFMPro is one of very few DFM tools that sits inside all four major CAD vendors' systems, where each platform vendor's own DFM checker is captive to its platform. [Synthesis]

Engineering Intelligence Stack Mapping

  1. Intent — requirements services plus XLM.AI requirement decomposition/compliance. Document-level, not design-intent capture. [Evidence] 7
  2. Knowledge — DFMPro's rule library + Rule Manager is a genuine, productised manufacturing-knowledge capture system ("capture in-house tribal knowledge, knowledge of retiring workforce, best practices" — encoded, by humans, as rules). The Cisco engagement proves the capture loop works in practice. [Evidence] 1514
  3. Reasoning — deterministic rule evaluation only. The probabilistic reasoning layer (violation filtering, warpage prediction) exists as a blog post, not a product. [Evidence] 16
  4. Execution — services muscle: 223k people, ASAP's 1,600 German automotive engineers, Volvo-scale delivery. This is what an ESP uniquely adds. [Evidence] 21
  5. Feedback — DFX Analytics collects check-result data, and auto-ignore collects reviewer dismissals — but no evidence any of it feeds a learning loop from measured production outcomes (defects, warpage, scrap). Consistent with every vendor in this study. [Evidence of absence] 13

Strengths

  • Owns the incumbent multi-CAD rules-DFM product for injection moulding — shipping since the Geometric era, embedded in CATIA/NX/Creo/SOLIDWORKS/ 3DEXPERIENCE. [Evidence] 10
  • Rule Manager + services model: customers' tribal knowledge is encoded as custom rules by HCL engineers (Cisco-verified) — a working knowledge-capture business. [Evidence] 14
  • Real automotive engineering scale in Europe: ASAP's 1,600 German engineers; Volvo strategic-supplier status; SDV analyst leadership. [Evidence] 218
  • Genuine (rented) GenAI platform with enterprise governance — AI Force — and a PLM-aware AI line (XLM.AI) that already does part search and process-plan generation over lifecycle data. [Evidence] 47
  • Unusual honesty: the DFMPro team publicly resists AI-washing its own product — a credibility asset in a hype-saturated market. [Evidence] 16

Weaknesses

  • No learned AI anywhere in the moulded-parts toolchain — DFMPro is pure rules; the AI augmentation layer is unbuilt strategy. [Evidence] 16
  • No simulation capability — no mould-flow, no structural solver; HCLTech cannot close the loop from DFM warning to physics validation without a partner (Moldflow, Moldex3D…). [Evidence of absence]
  • AI investment points away from mechanical — AI Force's pilots and modules are software/IT/financial-services; automotive strength (ASAP, SDV) is E/E-and-software-weighted. [Evidence] 48
  • Thin public automotive evidence for DFMPro — one logo (Ford Otosan), no automotive case study, no quantified moulding outcome. [Evidence of absence] 12
  • LLM layer rented (Azure OpenAI/Bedrock/Gemini) — like every vendor in this study. [Evidence] 46

Current Gaps (largely manual today)

Even with DFMPro deployed: choosing which violations matter (alert fatigue is the blog's own headline problem), predicting warpage/sink/cooling effects (needs simulation or learned models HCLTech doesn't have), translating tribal knowledge into rules (done manually by HCL engineers per engagement), and searching past part designs by geometry (XLM.AI's part search is metadata-level). Mould-flow interpretation and tooling design sit entirely outside HCLTech's product scope. [Synthesis] 167

Future Direction

  • The published DFM roadmap-in-spirit is the Feb 2026 blog's "sandwich architecture": rules as foundation, AI as violation filter/context layer (learning from waivers, mining ECOs/NCRs for candidate rules, supplier-specific tuning), human as decision layer. No dates, no committed features. [Evidence — strategy only] 16
  • AI Force 2.0 (Apr 2026) signals platform consolidation around agentic AI — if DFM ever gets an AI layer, this is the plumbing it would ride on. [Evidence + Inference] 5
  • Automotive ER&D expansion continues (ASAP integration, Volvo, Gothenburg CoE) — an ever-larger delivery channel into OEM mechanical programmes. [Evidence] 1
  • My read: HCLTech has the pieces (rules engine, CAD seats, agent platform, delivery arm) and has written down the design, but shows no urgency to build learned DFM — its incentive is billable configuration of the rules it already owns. [Inference]

Relevance to Automotive Moulded Parts

Capability Strength Notes
Plastic Part Design Services only Design delivery on client CAD; no product
Surface Design (Class A — styled, customer-visible show surfaces) Services only No public automotive evidence
CAD Automation Moderate DFMPro automates the checking, not the modelling
DFM Strongest — rules only Injection-moulding module: wall thickness, ribs, draft, undercuts, thin-steel; Rule Manager; multi-CAD; no AI11
Tool Design Weak Flags mould thin-steel conditions; no tooling product
Mould Flow None No solver, no partner product
Manufacturing Engineering Services Plus XLM.AI process-plan generation (document-level)7
Quality Services ASAP testing/validation (E/E-weighted)2
Engineering Knowledge Reuse Notable Rule Manager knowledge capture (Cisco-proven); XLM.AI part search & reusability engine — both non-geometric147

Critical finding. HCLTech is the mirror image of the platform vendors. They have solvers without services; HCLTech has services and a rules product without any solver or any learned AI. Its DFMPro blog states the field's honest truth in the vendor's own voice — rules are deterministic and irreplaceable, AI for moulding is potential, not product ("AI may be able to predict complex cooling-induced warpage"). For this study, HCLTech is the proof that rules-based moulding DFM is a viable, decades-old product category — and simultaneously proof that even its owner has not crossed into learned DFM. [Evidence + Synthesis] 16

Relevance to design-reuse retrieval. DFMPro validates the rules-DFM half of such a system (the checks are the same textbook rules; the commercial moat is CAD integration + rule configuration services). It also validates the retrieval gap: nothing at HCLTech searches historical parts by geometry — XLM.AI's "intelligent part search" is lifecycle-data search. [Synthesis] 7

Key Takeaways

  1. HCLTech is the ESP that productised rules-based DFM: DFMPro, ex-Geometric (acquired 2016), with a dedicated injection-moulding module inside all major automotive CAD systems. [Evidence] 1017
  2. DFMPro contains no learned AI — and HCLTech's own product team says rules are the bedrock; its moulding-AI examples are future-tense. [Evidence] 16
  3. The only learning-flavoured feature claim is auto-ignore ("learns from organizational actions") — undocumented as ML; treat as a feedback heuristic. [Evidence — thin] 16
  4. AI Force (Mar 2024; 2.0 Apr 2026) is real, rented-LLM GenAI — scoped to software/data/IT, not mechanical engineering. [Evidence] 45
  5. XLM.AI brings GenAI to PLM/ALM data (part search, process plans, BOM checks) — the closest AI to the moulded-parts workflow, but document-bound and showcased on pharma, not automotive. [Evidence] 7
  6. Verified moulded-parts customer evidence: Cisco (plastic + sheet-metal + die-cast rules, qualitative) and a 60% checking-time claim at Rockwell; automotive DFMPro evidence is a single logo (Ford Otosan). [Evidence] 141012
  7. Automotive scale is real but E/E-weighted: ASAP (~1,600 German engineers, Jul 2023, ~€251M) and Volvo Cars strategic-supplier status (Jun 2025). [Evidence] 231
  8. HCLTech has no mould-flow capability and no learning from production outcomes — consistent with every vendor studied. [Evidence of absence]
  9. For anyone pursuing moulding AI, HCLTech reads as the rules incumbent, a plausible white-label channel, and a potential customer — the opportunity lies in the intelligence layer HCL described in print but never built. [Inference] 16

References

Primary Sources — HCLTech

  • ASAP Group acquisition press release, Jul 12 2023 — https://www.hcltech.com/press-releases/hcltech-acquire-german-automotive-engineering-services-company-asap-group
  • Volvo Cars engineering-services press release, Jun 12 2025 — https://www.hcltech.com/press-releases/volvo-cars-and-hcltech-collaborate-drive-transformation-engineering-services
  • AI Force launch press release, Mar 5 2024 — https://www.hcltech.com/press-releases/hcltech-launches-ai-force-accelerate-time-value-software-development-and-engineering
  • AI Force 2.0 press release, Apr 1 2026 — https://www.hcltech.com/press-releases/hcltech-launches-ai-force-20-deliver-enterprise-grade-agentic-ai
  • AI Force platform page — https://www.hcltech.com/ai-force
  • XLM.AI product page — https://www.hcltech.com/xlm-ai
  • Automotive services page — https://www.hcltech.com/automotive-services
  • PLM / Digital Design & Manufacturing page — https://www.hcltech.com/product-lifecycle-management-plm

Primary Sources — DFMPro (HCLSoftware)

  • DFMPro site — https://dfmpro.com/
  • DFMPro for Injection Molding — https://dfmpro.com/manufacturing-processes/dfmpro-for-injection-molding/
  • DFMPro automotive industry page — https://dfmpro.com/industries/auto/
  • DFX Analytics — https://dfmpro.com/dfx-platform/dfx-analytics/
  • About DFMPro — https://dfmpro.com/about-dfmpro/
  • DFMPro brochure, Nov 2024 (Cisco testimonial, p.6) — https://cdn.dfmpro.com/wp-content/uploads/2024/11/New-DFMPro-brochure_Nov2024.pdf
  • DFMPro brochure v4.1, Jan 2026 — https://cdn.dfmpro.com/wp-content/uploads/2026/01/RGB_DFMPro-brochureV4.1.pdf
  • "Beyond the Hype: Why Rules-Based Systems Are Still the Bedrock of AI-Driven DFM," Rahul Rajadhyaksha, Feb 2 2026 — https://dfmpro.com/blog/beyond-the-hype-why-rules-based-systems-are-still-the-bedrock-of-ai-driven-dfm/

Secondary Sources

  • Reuters on ASAP deal value (~€251M / $280M), Jul 13 2023 — https://www.reuters.com/markets/deals/indias-hcltech-buy-german-automotive-services-firm-asap-280-mln-2023-07-13/
  • Wikipedia, HCLTech (Geometric acquisition Apr 2 2016; segments; headcount) — https://en.wikipedia.org/wiki/HCLTech

Appendix A — Timeline

Apr 2, 2016 — Geometric Ltd acquisition announced (DFMPro joins HCL) → Jul 12, 2023 — ASAP Group acquisition announced (closed later in 2023) → Mar 5, 2024 — AI Force launched (software-lifecycle GenAI, Azure OpenAI) → 2024–2025 — AI Force LLM integrations (Copilot, Gemini, Claude/Bedrock) → Nov 2024 — DFMPro brochure with Cisco plastic-parts testimonial → Jun 12, 2025 — Volvo Cars strategic engineering-supplier agreement → Feb 2, 2026 — "Beyond the Hype" rules-vs-AI DFM blog → Apr 1, 2026 — AI Force 2.0 (agentic, model-agnostic).

Appendix B — Glossary

  • ESP — engineering service provider: sells engineering labour and outcomes, unlike platform vendors who sell software.
  • ER&D / ERS — engineering and R&D services, HCLTech's segment name for that business.
  • DFMPro — HCL's CAD-integrated rules-based DFM checker (ex-Geometric).
  • Rule Manager — DFMPro component that turns an organisation's own design guidelines into automated checks.
  • DFX — "design for excellence": umbrella for DFM plus assembly, cost, service etc.; names DFMPro's Server/Analytics add-ons.
  • Thin steel condition — a part shape forcing a fragile thin section in the mould's steel; a tooling-risk DFM check.
  • NPI — new product introduction (the Cisco testimonial's context).
  • ECO / NCR — engineering change order / non-conformance report; the historical records the blog proposes mining for rules.
  • ALM / PLM — application / product lifecycle management systems (XLM.AI's operating ground).
  • SDV — software-defined vehicle.
  • Alert fatigue — engineers ignoring a checker that raises too many low-value violations; the problem DFMPro's proposed AI layer targets.

Appendix C — Notes (thinnest-evidence areas to revisit)

  1. Geometric deal mechanics — the April 2, 2016 announcement is Wikipedia-sourced here; the widely reported completion (~March 2017) and the carve-out of Geometric's 3DPLM joint-venture stake to Dassault Systèmes could not be verified against a live primary source in this pass.
  2. ASAP closing date and price — announcement PR verified directly; the Aug 31, 2023 closing (ASAP newsroom) and ~€251M price (Reuters) were confirmed only at search-snippet level.
  3. "Auto-ignore … learns" — single blog sentence; no product documentation found describing its mechanism. Could be ML, is more plausibly a stored dismissal list.
  4. Cisco testimonial dating — verified in the Nov 2024 brochure; an earlier research note's "Apr 2023 brochure" instance was not independently retrieved.
  5. "Patented GenAI platform" (AI Force) — no patent numbers located.
  6. DFMPro inside automotive ER&D engagements — no public evidence found that HCLTech deploys DFMPro within its own automotive services programmes (the Cisco engagement is the only documented product+services pairing); plausible but unproven.
  7. Rockwell 60% figure — vendor-hosted testimonial context; methodology unpublished.
  8. Open source / publications — searched at low depth only (session search budget); a deeper pass could yet surface Geometric-era papers.

Executive Summary

  • Who they are. HCLTech is one of the world's largest engineering-services companies: 223,000+ people in 60 countries, consolidated revenue of $13.8 billion for the twelve months ending March 2025, and a claim of working with 63 of the top 100 global engineering R&D spenders. [Evidence + Marketing] 12
  • Why they matter for moulded parts. Unusually for a services company, HCLTech owns a shipping DFM software product: DFMPro, inherited with the 2016 acquisition of India's Geometric Ltd. DFMPro is a CAD-integrated, rules-based manufacturability checker with a dedicated injection-moulding module, running inside CATIA V5, 3DEXPERIENCE, NX, Creo and SOLIDWORKS — the exact CAD systems automotive programmes use. [Evidence] 1017
  • AI maturity: real GenAI, wrong department. HCLTech's flagship AI platform AI Force (launched March 2024, built on Azure OpenAI) automates the software development lifecycle. Its PLM-facing AI, XLM.AI, applies GenAI to lifecycle documents and data (requirements, BOMs, test cases, part search). Neither touches mechanical part design or moulding. [Evidence] 47
  • The central honest finding. DFMPro — the one HCLTech product squarely in this study's domain — contains no learned AI. It is deterministic rules, and, remarkably, HCLTech says so itself: a February 2026 DFMPro blog post is titled "Beyond the Hype: Why Rules-Based Systems Are Still the Bedrock of AI-Driven DFM." Its injection-moulding AI examples are written in the future tense ("AI may be able to predict complex cooling-induced warpage"). An ESP has already productised exactly the rules-DFM layer of a design-reuse retrieval system — and has publicly mapped, but not built, the learned layer on top. [Evidence] 16

  1. Volvo Cars and HCLTech engineering-services press release, Jun 12 2025 — https://www.hcltech.com/press-releases/volvo-cars-and-hcltech-collaborate-drive-transformation-engineering-services ↩↩↩↩↩↩↩↩↩↩

  2. HCLTech to acquire ASAP Group press release, Jul 12 2023 — https://www.hcltech.com/press-releases/hcltech-acquire-german-automotive-engineering-services-company-asap-group ↩↩↩↩↩↩↩↩↩↩↩

  3. Reuters, "India's HCLTech to buy German automotive services firm ASAP for ~$280 mln," Jul 13 2023 — https://www.reuters.com/markets/deals/indias-hcltech-buy-german-automotive-services-firm-asap-280-mln-2023-07-13/ ↩↩

  4. HCLTech launches AI Force press release, Mar 5 2024 — https://www.hcltech.com/press-releases/hcltech-launches-ai-force-accelerate-time-value-software-development-and-engineering ↩↩↩↩↩↩↩↩↩↩↩↩↩↩

  5. HCLTech launches AI Force 2.0 press release, Apr 1 2026 — https://www.hcltech.com/press-releases/hcltech-launches-ai-force-20-deliver-enterprise-grade-agentic-ai ↩↩↩↩↩↩

  6. AI Force platform page (modules, LLM integrations, impact claims), accessed Aug 2026 — https://www.hcltech.com/ai-force ↩↩↩↩↩↩↩

  7. HCLTech XLM.AI product page (features incl. intelligent part search, process-plan generation; pharma case studies), accessed Aug 2026 — https://www.hcltech.com/xlm-ai ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

  8. HCLTech automotive services page (SDV/ISG analyst leaderships; body/chassis/powertrain services), accessed Aug 2026 — https://www.hcltech.com/automotive-services ↩↩↩↩↩

  9. HCLTech PLM / Digital Design and Manufacturing page, accessed Aug 2026 — https://www.hcltech.com/product-lifecycle-management-plm ↩↩↩

  10. DFMPro site (CAD platforms, testimonials incl. Rockwell 60%, Jabil, Motorola; 20/12/10/10% marketing metrics), accessed Aug 2026 — https://dfmpro.com/ ↩↩↩↩↩↩↩↩↩↩↩↩↩↩

  11. DFMPro for Injection Molding (wall thickness, ribs, draft, undercuts, thin-steel checks; custom rules), accessed Aug 2026 — https://dfmpro.com/manufacturing-processes/dfmpro-for-injection-molding/ ↩↩↩↩

  12. DFMPro automotive industry page (Ford Otosan logo; ECO cost claim; Automotive Plastics webinar), accessed Aug 2026 — https://dfmpro.com/industries/auto/ ↩↩↩↩

  13. DFX Analytics product page (dashboards; no ML described), accessed Aug 2026 — https://dfmpro.com/dfx-platform/dfx-analytics/ ↩↩↩

  14. HCL DFMPro brochure, Nov 2024 — Cisco Systems testimonial (Biren Kothari, NPI Technical Lead Mechanical): plastic/sheet-metal/die-cast rules digitised — https://cdn.dfmpro.com/wp-content/uploads/2024/11/New-DFMPro-brochure_Nov2024.pdf ↩↩↩↩↩↩↩

  15. HCL DFMPro brochure v4.1, Jan 2026 — Rule Manager, DFX Server, knowledge database, 15+ processes incl. Injection Moulding; zero AI mentions — https://cdn.dfmpro.com/wp-content/uploads/2026/01/RGB_DFMPro-brochureV4.1.pdf ↩↩↩↩↩

  16. R. Rajadhyaksha, "Beyond the Hype: Why Rules-Based Systems Are Still the Bedrock of AI-Driven DFM," DFMPro blog, Feb 2 2026 (published 2026-02-02 per page metadata) — https://dfmpro.com/blog/beyond-the-hype-why-rules-based-systems-are-still-the-bedrock-of-ai-driven-dfm/ ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

  17. Wikipedia, "HCLTech" (Geometric Ltd acquisition listed Apr 2 2016; segments; ~223,889 employees Jun 2026), accessed Aug 2026 — https://en.wikipedia.org/wiki/HCLTech ↩↩↩↩↩

  18. On distinguishing rules/optimization from learned AI — see the concept note What is Optimization technology. DFMPro is the clean case: deterministic geometric rule evaluation, honestly labelled as such by its vendor — the study's honest-AI test ("learns from data vs computes from equations") places it firmly on the computes-from-rules side. ↩