Where it all began: choosing Agilytic

1. The cultural foundation

Before Alex joined Agilytic in 2018 as a Data Scientist fresh out of Solvay, he received an offer from a larger consulting firm. What set Agilytic apart was its specialized focus on data science and something less tangible but equally (if not more) important: organizational culture.

The contrast was stark. After progressing through interview rounds at the big consulting firm, complete with formal handshakes and suits, Alex encountered a different atmosphere at Agilytic: we prioritize letting candidates meet their future colleagues before signing—an uncommon practice that proved decisive for him. Watching colleagues like Jérôme demonstrate sophisticated forecasting models revealed the level of technical expertise he would work alongside.

This emphasis on team composition over brand name reflects one of our core hiring principles: professionals join teams of people first, company names second. The ability to assess technical competence, cultural fit, and conflict resolution mechanisms before accepting an offer provides essential information that job descriptions cannot convey.

2. Projects that define impact

Throughout his years at Agilytic, Alex worked on a variety of projects showcasing the wide range of data science applications, including lead scoring and automation projects.

💯 Lead scoring

These projects focus on optimizing marketing efficiency by predicting which customers show the highest propensity for specific products. These models aggregate both internal company data and scraped publicly available information, particularly valuable in B2B contexts. They help organizations concentrate budgets where they generate maximum impact.

💵 Cost reduction through automation

One of Alex’s main projects involved transforming monthly reporting processes—that provided only retrospective insights—into daily reporting systems. This shift from monthly to daily visibility gave managers the agility to identify friction points and make immediate adjustments rather than waiting weeks for actionable information.

The entrepreneurial detour

1. A new adventure

After four years of building technical expertise in data science, Alex cofounded Leadix when a university friend approached him with increasingly technical questions about a new venture.

What began as informal consulting conversations culminated in an unexpected question: where would you look for a CTO? Alex's immediate answer was: suggesting alumni networks and specific individuals before saying "or I could do it". This marked the beginning of a two-and-a-half-year entrepreneurial chapter.

The experience reinforced Alex’s commitment to ownership: while Agilytic cultivates strong ownership culture, founding a company elevated this concept further. When something needs doing in a small startup team, it falls to the founders—whether it involves tedious administrative work or correcting others' mistakes. This often means unplanned late hours and weekend work, as early-stage ventures face existential risks where execution determines survival.

2. Returning where it all began

As Leadix grew to a team of 7 people, a divergence in vision with Alex's cofounder led to an amicable separation where he sold his shares. Rather than immediately pursuing the next opportunity, he deliberately took time to assess what he wanted:

• technically interesting work,

• recognition for personal contributions without the all-consuming intensity of startup founding,

• a culture emphasizing ownership and collaboration.

Listing these criteria made the answer obvious: since Agilytic already met all of them, it only made sense to come back. This decision reflects a mature understanding that exceptional opportunities are rare, and that constantly seeking something better can obscure the value of proven environments.

3. Evolution and continuity

Returning to Agilytic after two years, he noticed both consistency and growth.

The core culture, team energy, and collaborative spirit remained intact. But what had evolved was the methodology, refined through over 350 successful client projects. This accumulated experience allows faster, more assertive approaches to familiar problems.

We professionalized our documentation, onboarding processes, client engagement protocols, and project monitoring—all the while preserving our core culture.

4. Learning from missteps

Reflecting on his career trajectory, Alex identifies "careful what you wish for" as a critical lesson: focusing on rapid advancement and accumulating responsibilities, driven by status considerations and salary increases, isn’t a real career strategy. Additional responsibilities carry implications beyond LinkedIn titles.

His advice to his younger self: take your time. Careers span 40 to 50 years, and decisions made in the first five years don't determine everything that follows. People overestimate what they can accomplish in one year while underestimating what they can achieve in ten. Trajectory matters more than short-term objectives: rather than fixating on promotions by specific dates, focus on aligning your current path with your long-term direction.

This applies to decision-making too. Most decisions feel bigger than they are, and many have limited consequences or are reversible. Evaluate decisions by their potential impact and reversibility to gauge appropriate concern. Most career choices, including first jobs, fall into categories where mistakes are manageable and alternatives exist.

Alex’s commitments at Agilytic

1. Helping organizations navigate uncertainty

Organizations facing macroeconomic volatility and rapid technological change often ask: what should we do with data? Alex's guidance draws from Stoic philosophy: focus on what falls within your control. Macroeconomic conditions lie outside this sphere; operational processes, commercial efforts, and internal optimization do not.

During uncertain times, projects targeting cost reduction and operational optimization perform best. Automation, resource allocation optimization, and efficiency improvements become particularly valuable.

Additionally, uncertainty makes organizations risk-averse regarding vendors. Established consultancies with decade-long track records and hundreds of completed projects inspire more confidence than newcomers marketing AI expertise based solely on ChatGPT familiarity.

Smaller expert teams often deliver better value than large consulting firms that deploy numerous junior resources. Experienced practitioners who have solved similar problems multiple times provide greater guarantees regarding impact, timelines, and budget adherence.

2. Building data science capabilities

When building internal data science teams, technical competence is the baseline requirement. But abundant free resources and formal training programs make hard skill development accessible; the real differentiating factor lies in non-technical competencies.

For instance, client empathy—the capacity to understand problems before offering solutions—is a critical concept that often proves surprisingly difficult to implement. The gap between adequate and excellent data scientists often comes down to how well they comprehend stakeholder needs.

In this context, the user story framework from Agile methodology ("As [role], I need [function] to achieve [benefit]") provides a powerful structure. It encourages role-based thinking and clarifies desired outcomes versus functionality, avoiding premature solutions before understanding actual requirements.

3. Providing strategic guidance for data initiatives

When skeptical decision-makers question data investments, Alex's elevator pitch is direct: organizations accumulate competitive disadvantage with each passing day. The best time to plant a tree was 20 years ago; the second-best time is today. The same applies to AI and data capabilities.

The most effective approach inverts typical AI conversations. Rather than beginning with technology, gather stakeholders by department to discuss problems:

• what consumes excessive time,

• what prevents additional sales,

• what causes sleepless nights…

Only after mapping these pain points should conversations turn to whether AI or data science can provide solutions.

This problem-first, technology-second sequence ensures that data initiatives address actual organizational needs. Implementing solutions in search of problems is a common trap that undermines many well-intentioned projects.

Perspectives on the industry

Essential tools for modern data work

1. Call recording and transcription tools

These tools automatically generate meeting minutes, summaries, and action items. They now perform this work better than junior consultants did years ago, making their absence almost inexcusable.

2. Large Language Models

While ChatGPT and the likes often demonstrates superior pure performance, Copilot's integration within the Microsoft ecosystem provides security guarantees equivalent to SharePoint and Outlook. Organizations comfortable with Microsoft's security posture for confidential information can extend that trust to Copilot.

On the other hand, enterprise-grade solutions (including professional Perplexity licenses or custom internal AI systems) offer additional security layers and fine-tuning capabilities.

3. Friction-reduction tools to smooth workflows

These tools don't encourage laziness; they amplify human intelligence, allowing professionals to deliver higher quality work. For example, text expansion utilities like TextExpander allow rapid deployment of prompt libraries and standardized text blocks.

Building shared repositories of effective prompts for ChatGPT, Copilot, and other AI tools also creates reusable starting points for common tasks.

Agilytic's recent policy evolution reflects this reality. Previous restrictions on AI tools, particularly for code generation, gave way to recognition that productivity and intelligence gains justify broader access. However, this requires training on tool limitations and secure environments: failing to provide effective internal tools means employees may use public versions like ChatGPT, where confidential information pasted into prompts creates serious security vulnerabilities.

More on Alex

Creating a digital twin of Earth

Beyond data science, Alex is curious about ambitious concepts like creating a digital twin of Earth. This would enable testing policies, startup impacts, and macroeconomic scenarios, improving decision-making for global challenges including environmental sustainability.

In this project, the technical challenge lies not in creating complexity, but in combining numerous simple models into a holistic system. Just as individual data science models aggregate to optimize entire organizations, planetary-scale simulation would require interconnecting physical, chemical, and social models to represent complex emergent phenomena.

Book recs

Among his reading recommendations, Alex highlights Storytelling with Data as essential. The book's premise is straightforward: if clients cannot understand your visualizations, the fault lies with the creator, not the audience. Despite its apparent simplicity, applying these principles requires discipline. The book is worth reading multiple times to remember the basics: keep things simple and easy to understand.

The recurring theme throughout Alex's journey, from startup employee to founder to returning manager, centers on intentionality. Thoughtful assessment of needs and long-term trajectories consistently outperforms reactive decisions driven by external pressures or status markers. For data science professionals and organizational leaders, this measured approach provides a framework for navigating both technological acceleration and economic uncertainty.