Projects

01 / 08

Architect and Engineering Lead on a next-generation IT operations model centred on synthetic staff — goal-driven agentic AI systems that replace manual toil and reactive support with autonomous, context-aware execution. Synthetic agents emulate SREs, L1 support engineers, and security operators, operating within OPA-enforced guardrails while integrating with enterprise toolchains to execute real-world remediation actions.

• –Designed a multi-agent framework enforced with Nvidia NeMo guardrails where agents with multimodal awareness, semantic reasoning, and intent understanding proactively detect, diagnose, and remediate infrastructure issues without human initiation.
• –Implemented proprietary RBAC and OPA policy enforcement ensuring synthetic staff operate within defined boundaries and escalate to human engineers only for high-judgment edge cases.
• –Built the agentic integration layer connecting synthetic operators to enterprise monitoring, ITSM, and runbook automation platforms to close the loop on real-world actions across Networks, Cloud & Data Center, Orchestration Platforms, and End-User Services.
• –Delivered measurable reduction in L1 toil while preserving human control for high-stakes operational decisions — freeing engineers for resilience engineering and strategy.

02 / 08

Led enterprise transformation to modernise finance operations by enabling compliant e-invoicing across AP and AR ecosystems in line with European Union mandates. Delivered API wrappers that abstract legacy invoice workflows and enable seamless integration with platforms like Pagero, underpinned by a deep understanding of AP and AR processes. The solution was implemented across 2 different customers spread over Belgium and Poland, with plans in-place for release to more countries.

• –Designed API abstraction layers over legacy invoice workflows enabling compliant e-invoicing without disrupting existing ERP configurations.
• –Delivered event-driven, high-fidelity machine-to-machine integrations with Pagero and connected e-invoicing networks, secured via OAuth2.
• –Enabled real-time, interoperable data exchange between AP and AR systems, replacing batch-file workflows with fully digital pipelines.
• –Managed a smooth transition from traditional invoicing to compliant digital operations, measurably reducing invoice processing cycle times.

03 / 08

Architecting an enterprise Applicant Tracking System integrating with major job boards and enterprise HR platforms. Designed for zero-friction candidate sourcing across Naukri, LinkedIn, and internal channels, with deep ERP and CRM integration ensuring consistent candidate data across the full recruitment lifecycle.

• –Deployed AI-led career services for prospective candidates — covering role discovery, fit assessment, and guided application workflows that improve the candidate experience from first touchpoint to offer.
• –Integrated AI-driven skill assessments with advanced proctoring enabling structured, bias-resistant evaluation across technical and cognitive domains.
• –Designed to reduce end-to-end cycle time for a successful hire by half post-launch, through parallel pipeline processing and automated screening at each funnel stage.
• –Delivered advanced resume-matching and interview analytics increasing candidate pass-through rates by improving alignment between job requirements and shortlisted profiles.
• –Built smart job recommendation capabilities surfacing more suitable opportunities to candidates based on skills, experience, and trajectory — reducing misapplication rates and improving funnel quality.
• –Implemented a skills matrix management layer and candidate experience feedback loop providing continuous visibility into competency gaps and talent pool composition across the organisation.

04 / 08

Designed and led an enterprise agentic AI transformation, extending autonomous AI capabilities across HR, Finance, Legal, Marketing, and cross-functional operations. A multi-agent orchestration layer where domain-specific AI agents handle complex workflows with shared context, structured handoffs, and human-in-the-loop controls at every decision boundary — positioning AI as a true first-responder that not only synthesises information but executes actions across enterprise systems.

• –HR — Embedded an AI-first employee experience layer within internal portals, delivering policy-aware support across HR, Finance, GRC, Marketing, and CIS domains while automating leave and attendance management, insurance claims, and payroll queries (PF, payslips, Form-16). Strengthened alumni engagement and offboarding journeys through guided workflows and automated nudges, reducing support ticket volumes by ~80%.
• –Talent Management — Delivered an AI-driven talent ecosystem spanning career discovery, hiring, and internal talent operations, including AI career assistants, recruiter dashboards, and intelligent interview workflows. Improved candidate engagement and evaluation consistency, positioning hiring as a data-driven, continuously learning system rather than a manual, fragmented process.
• –Finance — Established an AI-powered first-responder layer within enterprise email using Microsoft Graph integrations to automate vendor communications for accounts payable — processing and verifying invoices against GRNs, POs, and PRs, then generating accurate, context-aware responses to vendor queries. Handling ~600 emails per month, the system eliminates workload equivalent to a four-person team.
• –Marketing — Established an AI-driven website assistant that transforms static digital presence into an interactive, intent-aware engagement channel capable of answering queries on services, partners, leadership, and brand positioning while orchestrating downstream actions such as meeting scheduling. Captures conversational context to enhance lead qualification and shift from passive content delivery to conversion-oriented engagement.
• –Legal — Extended the platform into legal operations with AI-enabled contract lifecycle management, including intelligent redlining and contract review embedded directly within Microsoft Word via add-ins. Enabled template-driven NDA generation, validation, and compliance checks using intent understanding and reasoning — reducing turnaround time while ensuring policy alignment across legal workflows.
• –Cross-Functional Platform — Unified all capabilities under a platform-centric agentic architecture built on semantic relationships, multimodal awareness, reasoning, and intent understanding. Autonomous yet controlled operations across functions allow human teams to focus on strategy, judgment, and high-impact decision-making.

05 / 08

Led enterprise AI transformation for a major financial services firm across two parallel programmes — one spanning the corporate tax and financial support divisions, the other covering audit and consulting. The engagement began with structured requirements workshopping with business partners in each division, converging on a common integration plane that enabled AI-driven workflows across otherwise siloed functions.

• –Workshopped requirements directly with business partners in the corporate tax, financial support, audit, and consulting divisions — translating domain-specific workflows into structured AI use cases with clear boundaries and success criteria.
• –Designed a common integration plane that unified disparate systems across both programmes, enabling AI agents and workflows to operate across divisional boundaries without bespoke point-to-point integrations.
• –Delivered AI-driven workflow automation across the corporate tax and financial support divisions — covering document processing, query handling, and decision-support use cases previously requiring significant manual effort.
• –Extended the platform to audit and consulting divisions, enabling AI-assisted engagement workflows, intelligent document review, and context-aware client support grounded in firm-specific knowledge.

06 / 08

Part of the three-member founding team that built Microland's AI Centre of Excellence from the ground up — the practice that now underpins every intelligent product the company ships. The team was responsible for reference architecture, model governance framework, and an internal accelerator library that reduced AI project ramp-up from six weeks to ten days.

• –Co-defined the reference architecture for AI initiatives now used as the baseline across all Microland AI projects.
• –Co-established the model governance framework covering explainability, fairness, and production monitoring standards.
• –Built the internal accelerator library reducing baseline AI project setup from six weeks to ten days.
• –Stood up the RAG and LLM tooling standards that underpin Microland's production intelligent products.

07 / 08

Led architecture and delivery of 5G network orchestration platforms for telecom clients, deploying and managing containerised network functions across bare metal, Kubernetes, OpenShift, and cloud-native telecom infrastructure. Covered the full stack from underpinning bare metal provisioning through to VNF lifecycle management — including fully disconnected, air-gapped environments with strict network isolation requirements.

• –Deployed and configured bare metal infrastructure underpinning 5G workloads, and managed OpenShift clusters across IPI, UPI, and ZTP deployment modes to match varying site automation and connectivity constraints.
• –Set up disconnected, air-gapped OpenShift clusters for isolated network environments — including mirroring container registries, configuring offline operator catalogs, and managing disconnected update flows.
• –Delivered Rancher-based multi-cluster management enabling unified operations across geographically distributed 5G infrastructure and bare metal sites with a single control plane.
• –Implemented Robin storage and data management for stateful telecom workloads, and built automated VNF lifecycle management pipelines covering deployment, horizontal scaling, health monitoring, and graceful decommissioning.

08 / 08

Built NLP and machine learning-based automation solutions for enterprise use cases, leveraging RASA and Microsoft LUIS to deliver intent-driven automation across customer service, internal helpdesk, and process automation workflows. Contributed to the early ML and NLP foundations that later underpinned enterprise AI products.

• –Developed intent classification models using RASA and Microsoft LUIS achieving high accuracy across domain-specific enterprise vocabularies with ambiguous, real-world support language.
• –Built NLU pipelines handling entity extraction, context management, and multi-turn dialogue flow for conversational automation use cases.
• –Delivered production NLP automation across customer service and internal helpdesk channels, reducing manual handling time for routine query types.
• –Established NLP tooling conventions and reusable pipeline components later adopted as standards in Microland's AI Centre of Excellence.