Backup software usually enters the room after the interesting decisions have been made. The CRM is chosen, the workflows are built, the employees are trained. Then somebody asks a gloomy question: what happens when a person, an integration or an attacker mangles the data? Odaseva exists because, in 2012, Schneider Electric asked that question at a scale that made ordinary answers look flimsy.
The energy-management giant was replacing more than 100 legacy systems with a single Salesforce organization. Its new estate held eight terabytes of data, 12 terabytes of files and 1,200 custom objects across three Salesforce clouds. Some 45,000 employees and 700,000 partners depended on it. Existing products on AppExchange and MuleSoft could not satisfy the protection requirements. Schneider asked Salesforce architect Sovan Bin to study how cloud data could receive the mature controls the company had expected on premises.
Bin, described by Odaseva as the first Salesforce Certified Technical Architect in Europe, built a backup-and-restore system with Salesforce and Schneider Electric. The hard customer was not a distraction from the startup idea. It was the startup idea. Odaseva grew around the conviction that a global Salesforce estate is less like a tidy address book and more like a living industrial system - full of relationships, custom schemas, API limits, residency rules and unforgiving recovery objectives.
The product is the restore, not the copy
This distinction is the spine of Odaseva's pitch. Creating a copy is only the opening move. A usable recovery plan must detect what changed, preserve relationships between records, files and metadata, respect the order in which Salesforce accepts data, and return the business to a known state. It also has to avoid consuming so much API capacity that the protection system harms the production system it is meant to save.
One nonprofit customer illustrates the gap. Australian Network on Disability previously received Salesforce's weekly email saying an export had run. That was a backup ritual, but not an assured route home. The organization wanted to protect its full environment before releases and roll it back if code failed. Odaseva's case study says backups now run every four hours across more than 950 objects and millions of records while using no Bulk API calls and just 0.17 percent of available REST API calls.
“I can now backup the entire environment before I put a release out. If it fails, I can restore everything back.”Ari Perala, Australian Network on Disability
What failed first was often not a server. It was confidence in the old process. At Autism CRC, a previous protection vendor did not match the CRM team's needs; the organization wanted restore choices ranging from a single record to an entire Salesforce object at a chosen point in time. At ENGIE's Belgian business, moving customer operations into Salesforce created a demand for the same degree of control the utility expected from on-premises data. For Rémy Cointreau, a direct-to-consumer push made CRM data too central to leave behind a generic contingency plan.
A Swiss Army knife for one very large cloud
Odaseva expanded outward from recovery. Its security suite includes backup and restore, archiving, encryption, masking, privacy workflows and long-lived data trails. Its management suite covers sandbox seeding, replication, automation, org management and cross-org insights. The pieces solve related problems: move stale records out of expensive production storage; give developers realistic but masked test data; replicate information to another system; prove who changed a field; govern several Salesforce organizations without treating each one as an island.
That breadth is the competitive argument. Own, Salesforce Backup, Gearset, Spanning, GRAX, Druva, Veeam and others all occupy part of the landscape. Some are easier for a small administrator to buy. Some protect many SaaS applications instead of specializing in Salesforce. Gearset is closely associated with DevOps. Odaseva stakes out the uncomfortable end: enormous data models, many orgs, strict sovereignty rules, auditors at the table and architects who speak fluently about recovery points and API ceilings.
The expertise is unusually literal. Bin spent six years at Salesforce and helped lead its Paris architecture team. The founding and executive group includes other enterprise-software, growth and legal operators. Salesforce Ventures invested, while Salesforce remains a central partner. That closeness is a strength for customers committed to Salesforce and a concentration risk for anyone seeking a broad, application-neutral backup layer.
The backup copy gets a second job
The most consequential change in Odaseva's thinking is visible in Data Edge, introduced as the anchor of its 2025 product announcements. Instead of treating archived and protected data as cold insurance, Data Edge is designed to make it queryable for analytics, AI and continuity. It can unify information from Salesforce, ServiceNow and Microsoft Dynamics, connect governed CRM data to Agentforce, ChatGPT, Claude and Gemini, and feed Snowflake or Databricks through zero-copy architecture.
In company performance testing with a large financial institution, Odaseva reports 500 million SOQL records per minute and speeds from 300 to 1,800 times Salesforce Bulk API 2.0. It also advertises 99.99 percent availability with automatic API failover. Those are vendor tests and promises, not universal outcomes, but they reveal the strategy: the company wants to graduate from insurance policy to data control plane.
The logic is timely. AI agents are only useful when they can reach trustworthy context, but years of customer history may sit in archived objects, retired orgs or separate CRMs. Odaseva now talks about “data reversibility” - retiring a costly legacy org without discarding its information, then making that governed history available to analytics or AI. The same copy that lowers recovery risk can, in theory, lower storage duplication and widen the memory of an agent.
What it costs - and what a buyer can copy
Odaseva does not publish enterprise list pricing. Its smaller Professional Edition is priced per user and starts at $5,000 a year, including daily protection of data, files and metadata, restore capabilities and activity analytics. The larger platform is sold by quote, reflecting volume, products, service levels, residency and implementation complexity. Managed Backup Services are offered without an additional fee for eligible backup customers, according to the company.
The broader cost story is the capital behind the build. Odaseva raised $2.3 million in seed funding, $11.7 million in a 2019 Series A, roughly $25 million in a 2020 Series B, and $54 million in a 2024 Series C led by Silver Lake Waterman. Total funding exceeds $90 million. The latest round funded product development, zero-trust security and geographic expansion. This is not a weekend utility with a prettier dashboard; it is a specialist vendor taking on engineering and compliance obligations that large institutions expect to inspect.
The playbook worth borrowing
- Begin with the recovery objective, then work backward to backup frequency, retention and architecture.
- Test restores before an incident, including relationships, metadata, files and permission boundaries.
- Treat API capacity as a budget. Protection should not throttle the live business.
- Use one governed copy for several jobs where sensible: recovery, audit, masked testing, analytics and AI.
- Pick the vendor for the hardest credible failure, not the easiest demo.
The approach has boundaries. A small Salesforce organization needing cheap, set-and-forget daily backup may not benefit from the enterprise machinery; simpler products and native tools can cost less and demand less implementation. A company with little Salesforce exposure may prefer a vendor that protects many SaaS platforms with one policy. Data Edge's cross-CRM direction broadens the addressable market, but Odaseva's deepest proof still comes from Salesforce.
It also depends on organizational discipline. Granular controls do not replace a tested incident plan. Data residency features do not settle every legal interpretation. Fast query demonstrations do not guarantee identical performance against every schema. The platform works best when architects, security leaders, compliance teams and business owners agree on what must be recoverable, how quickly, in which jurisdiction and by whom.
The boring layer becomes the strategic one
Odaseva's best achievement may be making the backup copy interesting without forgetting why it exists. The company now protects more than 12 petabytes for hundreds of organizations and more than 100 million Salesforce users. It has gathered credentials for healthcare, government, automotive and Australian public-sector work, received FedRAMP Ready status in 2025 and joined France's French Tech 120 program.
But the founding scene remains the useful one: a huge customer tried to move its operational heart into the cloud and discovered that the safety system had not kept up. Odaseva built at the edge of what Salesforce could comfortably manage. Fourteen years later, it is betting that the edge has become the center - because resilience, sovereignty and AI all begin with knowing where the data is, who can touch it and whether you can get it back.