In the complex landscape of humanitarian crises, effective analysis increasingly depends on bridging global analytical frameworks with local understanding. For analysts like Annesha Mahanta, this means going beyond data to engage with local literature, cultural narratives, and community perspectives that provide essential context to numbers and trends. This integration of local knowledge transforms humanitarian analysis from a distant technical exercise into a grounded, relevant tool that reflects the lived realities of crisis-affected populations. Her approach involves deliberately incorporating local perspectives, being attentive to power dynamics, and understanding the historical and cultural contexts that shape both vulnerabilities and resilience. While tools like GANNET have revolutionized data processing capabilities, it is this human capacity to interpret information through culturally sensitive lenses that gives humanitarian analysis its transformative power.
In this third contribution to our Human Algorithm series, we interviewed Annesha Mahanta, an analyst at Data Friendly Space (DFS). Through her practice of reading local literature alongside technical reports and her commitment to amplifying community perspectives, Annesha demonstrates how the most effective humanitarian analysis happens when global analytical frameworks are infused with local knowledge. With her, we discussed the incorporation of local context and knowledge, the human costs of analyzing crises, what skills are needed as a humanitarian analyst and how to maintain these.
In what ways do you incorporate local context, culture, and social dynamics into your analysis?
Incorporating local context, culture, and social dynamics is absolutely essential to producing meaningful humanitarian analysis. One way I do this is by reading not just reports or assessments, but also local literature (fiction, essays, or even journalism) to better understand the lived realities, histories, and narratives of affected communities. This helps me go beyond numbers and grasp how people interpret crises, resilience, or displacement on their terms.
From an analytical standpoint, I ensure that secondary data is triangulated with local perspectives, whether from community-based organizations, local media, or qualitative findings from key informant interviews. I pay close attention to power dynamics, social hierarchies, gender norms, and historical grievances that can affect both vulnerabilities and access to services. Ultimately, embedding local context is what makes analysis useful, respectful, and grounded in the actual needs and capacities of crisis-affected populations.
How do you handle the emotional or psychological toll of analyzing distressing situations in crisis zones?
It’s definitely not easy analysing situations involving displacement, violence, or protection risks can be emotionally heavy. Over the years, I’ve learned to approach my work with both empathy and intentional boundaries. I remind myself that the purpose of my analysis is to amplify voices, highlight needs, and push for better-informed, timely responses. That sense of purpose helps ground me.
Practically, I try to build in space for reflection. Working remotely and across multiple crises, I also make sure to maintain routines that help me disconnect, whether it’s reading books (I am an avid reader), travelling to new places, going for hikes, or even just turning off notifications.
There’s also a collective resilience that comes from working with other analysts and broader teams who care deeply about getting it right. We support each other, and that solidarity makes a real difference.
What specific skills do you believe are essential for succeeding in humanitarian analysis?
Humanitarian analysis requires a blend of technical, contextual, and interpersonal skills. Strong analytical and critical thinking skills are essential being able to sift through vast amounts of fragmented data, identify trends, gaps, and triangulate findings is core to the role. Tools like GANNET support this, but the human judgment behind them is crucial. Contextual sensitivity is key to understanding the political, cultural, and historical nuances of a crisis helps ensure the analysis is both relevant and ethical. I’ve found that soft skills like empathy, curiosity, and open-mindedness are just as important as technical ones. I also believe adaptability is non-negotiable. In humanitarian settings, priorities can shift quickly, and being able to pivot, remain calm under pressure, and still produce clear, structured outputs makes a big difference.
What are the most useful resources or platforms for continuous learning in humanitarian analysis?
For continuous learning in humanitarian analysis, I rely on a mix of formal training platforms and real-time practice. The Kaya platform has been particularly useful, with its wide range of self-paced courses on needs assessments, protection analysis, and coordination, which helps reinforce both foundational and advanced skills. I also follow courses offered by the Harvard Humanitarian Initiative, especially those exploring the intersection of data, ethics, and crisis response—they offer great insight into the evolving nature of humanitarian work.
How has AI/GANNET changed the way you approach data analysis in humanitarian contexts?
GANNET has significantly changed how I approach analysis. What used to take hours of manual review and extraction can now be done in minutes. GANNET rapidly processes vast volumes of information from diverse sources media reports, humanitarian updates, assessments, and structures it in a way that allows me to focus more on actual analysis rather than spending the bulk of my time cleaning or organizing data. It’s especially valuable during sudden-onset crises, where timely insights are critical. Instead of being overwhelmed by the information flow, I can dive straight into identifying trends, gaps, and risks.
What potential does AI bring to traditional humanitarian methods and strategies?
AI systems are advancing to better understand local languages, dialects, and cultural nuances, which would make information gathering more inclusive and contextually accurate, but this is still just the start of such integrations. Another promising area is predictive analytics using AI to anticipate displacement, food insecurity, or protection risks before they escalate. That could transform how we prepare and respond. We’re also just beginning to tap into AI’s potential for early warning systems and sentiment analysis from social media or local news to detect emerging crises.
